Nuffield Council on Bioethics / Home

Ethics of Research involving animals

The validity of animal models used in pharmaceutical research

8.37 We have described why and how animals are used in pharmaceutical research and have illustrated with several examples the range of welfare implications that they may experience. Many people who are concerned about animal suffering are critical of the permissibility of animal research on ethical grounds. However, there are critics who also object to the use of animals in pharmaceutical research on scientific grounds. They question the transferability and predictability of data obtained from animals, and its reliability for the accurate assessment of the safety of new therapeutic interventions, as shown by the following respondents to the Consultation:

‘…animal experimentation is positively harmful to human health… [It] does not provide information that is relevant to human medicine because the data cannot be transferred to humans with any degree of reliability. In fact, studies of the predictability of animal experiments consistently show them to be worse than random guesswork… Adverse drug reactions are the fourth leading cause of death in the Western world, killing over 100,000 individuals every year in the US alone. Clearly, the animal tests are failing to protect people.’
Animal Aid
‘…claims that animal experiments have instilled a misplaced sense of the relative danger of a drug are supported by the incidences of false negatives and false positives known to be attached to such tests.’
Cris Iles-Wright

8.38 We have shown above that producing a new medicine is a lengthy and complex process, and that decisions on the compounds that should proceed to the next stage are taken using a wide range of information. Tests on animals play a vital role, but they are not the only source of information that is used to determine safety and efficacy (see Figure 8.3). Some critics of animal research and testing tend to attribute any problems with the final product solely to the use of animal testing. We consider the general question of whether or not animals are useful models for humans in medical research in paragraphs 10.27–10.32).
Systematic limitations faced by any modelling approach are addressed in paragraphs 10.33–10.36), and the findings of scientific reviews on the critical evaluation of research involving animals are discussed in paragraphs 10.37-10.43).

8.39 We observe that claims that animal research is failing to protect people from adverse drug reactions (ADRs) need to be treated with some caution. ADRs43 have a number of causes. Many of these are avoidable, for example where they arise from prescription errors, where people have been given or have taken the wrong medicine, or from interactions between different medicines taken simultaneously. In 2004, researchers conducting the largest prospective analysis in the UK of ADRs as a cause of admission to hospital found that more than 70% were avoidable and could have been predicted by taking into account pharmacological properties of the medicines involved.44 While ADRs may be the direct result of administration of one specific medicine, the question remains whether this is proof of the failure of the animal model (or any other model) involved in the development process, or a methodological problem. As we have said, phases I–II of human clinical trials in the development of a medicine include up to 5,000 patients to monitor efficacy and safety. If severe ADRs occur during these trials, the development of the medicine is not usually taken further. However, ADRs may occur at very low statistical frequencies, for example 1 in 10,000, and hence may not be revealed at this stage (see paragraphs 10.33 and 10.1). In making inferences about the occurrence of ADRs, and the role that animal research plays, it is therefore unhelpful to generalise. ADRs can occur for a number of reasons and could, in principle, also be caused by a medicine that, hypothetically, had been developed without the use of animals.

8.40 Some also argue that the withdrawal of medicines from the market is indicative of the fact that animal research does not help to prevent ineffective or harmful medicines being used by humans.45 In the UK, the Medicines and Healthcare products Regulatory Agency (MHRA) monitors whether medicines on the market meet the appropriate standards of safety, quality and effectiveness. When there is sufficient evidence to suggest that the risk of taking a medicine outweighs its benefit to patients, the Committee on Safety of Medicines (CSM) and MHRA take appropriate regulatory action to protect the health of patients, and may initiate steps to withdraw medicines from use. Between 1995 and 2005, 18 medicines were withdrawn from the UK market by companies or by the Licensing Authority on grounds of safety (see Box 8.6). A study conducted in 1994 on medicines withdrawn between 1961 and 1992 concluded that in the UK, 49 were taken off the market (see Box 8.7). These withdrawals were mainly due to inadequate evidence of efficacy in widespread clinical use, loss of therapeutic interest or poor market performance. To what extent the withdrawal of medicines can be attributed exclusively, or in part, to the use of animals in research would need to be assessed in individual cases (see paragraphs 10.27–10.43).46

Box 8.6: Medicines withdrawn in the UK
Box 8.7: Medicines withdrawn from the market*

8.41 Lastly, animal research is also undertaken by the pharmaceutical predictive capacity of data obtained from animal and researchers seek to identify how results from different order to develop better predictions of how the medicine excreted in the human body (see Boxes 8.8 and 9.4).

Box 8.8: Example of animal research undertaken to improve the predictability of pharmacokinetic data Aviles P, Pateman A, San Roman R et al. (2001) Animal pharmacokinetics and interspecies scaling of sordarin derivatives following intravenous administration Anitmicrob Agents Chemother 45: 2787–92.* The aim of this research was to compare the pharmacokinetics of a group of synthetic antifungal agents against reference compounds in different animal species and to assess whether human pharmacokinetics could be reliably predicted from this information. The antifungal agents that were used belong to a new group of synthetic chemicals called sordarin derivatives which have been shown to prevent the growth of fungal pathogens. Opportunistic fungal pathogens remain a common cause of death in immunocompromised patients, such as those with HIV/AIDS or those receiving chemotherapy or immunosuppressive therapy. The study used Cynomolgus monkeys, rats, mice and rabbits. Gunn rats, which have impaired liver function, were used to assess the processing of sordarin derivatives when they pass through the liver. A representative sordarin derivative was administered intravenously to animals. In mice, this was achieved by puncture of the tail vein. The compound was administered through tubes inserted into the jugular veins of rats and into marginal ear veins of rabbits. In monkeys, administration was performed via the cephalic vein. Each compound was administered once. In mice, blood samples were taken by cardiac puncture using a needle at eight intervals after administration. Three mice were euthanised by cervical dislocation at each sampling point. Samples of rat blood were taken from the end of the tail. Rabbits were sampled using a tube placed in the central artery of the ear. Samples of monkey blood were obtained from the posterior of the animals by direct venepuncture using a needle at ten intervals after administration. Blood samples were allowed to clot and then centrifuged to separate the serum (the clear yellowish fluid obtained upon separating whole blood into its solid and liquid components). Data obtained from the serum were used to construct concentration-time curves to evaluate the effectiveness of the substance on the growth of fungal protein. Comparisons between the species used were made and assessed against data from previous studies. The researchers concluded that integrating the pharmacokinetics data from the different animals used would result in better predictions for the way the sordarin derivatives are metabolised in humans and therefore contribute to the study design of initial clinical trials in humans. * This is an example of animal research that has been published in a peer-reviewed journal. Details relate to this specific example and should not be taken to represent a ‘typical’ animal experiment. It is important to note that individually published experiments usually form one part of a continuing area of research, and the significance of the results may therefore be difficult to interpret.

Footnotes

43 Edwards and Aronson define an ADR as ‘an appreciably harmful or unpleasant reaction, resulting from an intervention related to the use of a medicinal product, which predicts hazard from future administration and warrants prevention or specific treatment, or alteration of the dosage regimen, or withdrawal of the product.’ See Edwards IR and Aronson JK (2000) Adverse drug reactions: definitions, diagnosis, and management Lancet 356: 1255–9.
44 The researchers, using Edwards and Aronson’s definition of ADRs (see previous footnote), sought to ascertain the burden of
ADRs though a prospective analysis of hospital admissions to two large general hospitals in the UK. Every patient aged over 16 years who was admitted to these hospitals (18,820 patients) over a six month period was assessed to determine if the admission had been caused by an ADR. It was found that 1,225 admissions were related to ADRs (equalling 6.5%, which is consistent with an estimate of 5% based on pooled data from several studies worldwide). Three types of avoidability were assessed: definitely avoidable (7-10%: the ADR was due to treatment inconsistent with present day knowledge of good medical practice), possibly avoidable (60-66%: the ADR could have been avoided by an effort exceeding the obligatory demands of present day knowledge of good medical practice) and unavoidable (25-30%: the could not have been avoided by any reasonable means). See Pirmohamed M, James S, Meakin S et al. (2004) Adverse drug reactions as cause of admission
to hospital: prospective analysis of 18 820 patients BMJl 329: 15–9. See also Waller P and Rawlins P A User’s Guide to the Safety of Medicines, available at: http://www.dsru.org/pat_guide_1.html. Accessed on: 2 May 2005; Kohn LT, Corrigan JM
and Donaldson MS (Editors) (2000) To Err is Human: Building A Safer Health System, available at:
http://www.iom.edu/report.asp?id=5575. Accessed on: 26 Apr 2005.
45 See BUAV Don’t we need animal experiments to make sure drugs are safe for humans?, in Frequently asked questions
about vivisection, available at: http://www.buav.org/faqs.html. Accessed on: 2 May 2005.