Category Archives: OPINIONS

Read-across for Hazard Assessment: The Ugly Duckling is Growing Up

Wera Teubner and Robert Landsiedel

Increasing use of read-across in integrated approaches for the testing and assessment
of chemical hazards will ensure that it eventually matures into a beautiful swan

In the hazard assessment of chemicals, read-across describes a technique used to predict physicochemical, ecotoxicological and toxicological endpoints. If it is performed on several substances at a time, it is called ‘category formation’. Read-across is based on the experience that similar chemicals exhibit similar properties — with the crucial issue of knowing which properties determine similarity for a given endpoint. In this aspect, it is a relative of the quantitative/ qualitative structure–activity relationship (QSAR), and was sometimes simply termed ‘expert judgement’. The idea of the read-across concept being an ‘ugly duckling’ has mostly arisen from the difficulty in verifying the plausibility of its findings without actually performing the experimental studies. The Read-Across Assessment Framework (RAAF),1 published in May 2015, states that “Under REACH, any read-across approach must be based on structural similarity between the source and target substances”. However, the limited verification of readacross, and especially the limitations of the use of the read-across approach only to structural similarities, reflect a state of infancy that needs to be nurtured toward maturity in order to reap its maximum

When, in 1959, William Russell and Rex Burch published The Principles of Humane  Experimental Technique,2 calling for the replacement, refinement  and reduction of animal testing, a major focus was the quality of animal testing and the criticism that
poor planning and experimental techniques resulted in animal studies of limited value, and consequently in more testing than should have been needed. With the introduction of Good Laboratory Practice and of Organisation for Economic Co-operation and Development (OECD) test guidelines (TGs) and animal welfare policies, the quality of animal data has become much less of a problem, and refinement has considerably improved. The improvement of cell culture, tissue culture and molecular biology technology kindled the hope for replacement. Meanwhile, standalone in vitro methods (e.g. for skin and eye irritation) or batteries of tests (e.g. for skin sensitisation) can address local toxicity. Likewise, methods to address specific early effects or mechanisms, such as genotoxicity or oestrogenic activity, are available.

A major challenge today is the prediction of complex toxicological effects such as systemic  and developmental toxicity. Large research programmes, e.g. ToxCast or SEURAT, aim to meet this challenge.3,4 Any new approach to complex toxicological effects combines various methods (in silico, in vitro and in vivo) in a testing battery or strategy.5,6 These approaches use mechanistic information, and are constructed according to (putative) adverse-outcome pathways (AOPs).7 Such information is, of course, also useful in supporting the read-across of apical toxic effects of different chemicals. Read-across can actually become a successful part of many integrated approaches for testing and assessment (IATAs).

Traditionally, chemicals are considered candidates for read-across, if they share structural similarity or are metabolically or spontaneously transformed to common products. It is assumed that structural similarity will result in a common mode-of-action. When assessing wanted pharmacological activities or unwanted toxicological hazards in research and development, applying read-across is already possible when the substance in question still only exists on paper. High-quality predictions are valuable for success in product development. At some point, the predicted effects are determined experimentally for promising candidates, and it is at this point that the consequences of poor read-across hit back. Again, identifying the correct similarity between read-across source and target chemicals is crucial.
Figure 1
The ‘ugly duckling’ characteristics of read-across (Figure 1) originate from areas in which it is used as a quick (and cheap) means to generate hazard information, either to fulfil regulatory data requirements, or to identify and list substances allegedly of very high concern (no reference given here, since this PiLAS is not a pillory). It also may originate from the idea that any information is better than no information in situations where there is no budget, or when animal testing is simply out of the question. Global efforts to identify and substitute hazardous chemicals can only succeed, if so-called ‘regrettable  substitutions’ can be avoided. Neither overestimation nor  underestimation of hazards by read-across is helpful in this context. Actually, it takes a wide range of thorough considerations to perform a robust and meaningful read-across — and these need to be documented.  To toxicologists with long experience in their respective chemical space, similarity may seem so obvious that their read-across justifications are rather frustrating to comprehend.

The application of read-across and the related category approach received a boost when  the European Union (EU) introduced the REACH programme in 2006. The REACH legislation (EC Regulation 1907/2006)8 requires the hazard characterisation of all chemicals marketed in the EU, with actual data requirements dependent on the production and import tonnage and the use conditions. With the estimation that more than 20,000 chemicals would need to be assessed, the legislation needed to include provisions to use animal testing only as a last resort. The obligation of the European Chemicals Agency (ECHA) to report on the status of the implementation and use of non-animal test methods and testing strategies is actually laid down in Article 117(3) of the legislation. As of 1 October 2013, dossiers for 8,729 substances have been submitted to the ECHA. A readacross or category approach was used in up to 75% of analysed dossiers for at least one endpoint.9

Considering the huge number of chemicals that were to be registered within the short period of eight years, the REACH legislation introduced a previously mostly-unknown component to chemical legislation. It was proposed that acceptance of registration, if appropriate, would be granted after automated dossier screening. Any scientific review of toxicological data would then be performed at a later stage, and this review would have to be conducted for at least 5% of the registered substances. With this procedure, the opportunity for an upfront discussion on the data requirements and suitability to support a read-across approach is in no way considered. This registration strategy has the advantage of speed and a certainty of meeting submission deadlines, but  he disadvantage of uncertainty with regard to follow-up activities, the latter arising from the possibility that the read-across assessment might be judged to be deficient and the decision would then be made that the target substance must be tested.

Both challenges and improvements to read-across approaches have been triggered by cases where apparently small changes in structures resulted in vast changes of the hazard properties (so called ‘activity cliffs’). The most prominent examples originate from differences in the interactions of substances with enzymes and receptors. The substances 2-acetylamino fluorene (2-AAF) and 4-acetylaminofluorene (4-AAF) are structurally very similar. As well as being a bladder carcinogen, 2-AAF is a strong liver enzyme inducer, leading in long-term studies to liver tumours. However, 4-AAF only slightly induces liver enzymes and does not induce the formation of liver tumours.12 Enantiomers of 1-hydroxyethylpyrene are activated to mutagenic sulphates by different sulphotransferases, 13 and the enantiomers of Carvone smell of caraway or spearmint,14 to name but two examples. When looking at the two-dimensional description of a chemical only (e.g. SMARTS pattern or Tanimoto score), stereoisomers appear identical, but three-dimensional structure modelling for receptorbinding simulation can differentiate stereoisomers. Regardless, stereo-isomeric and regio-isomeric differences of molecules appear to be small alterations, as compared to the changes usually bridged by readacross (e.g. homologous series). It is important to know which aspect of similarity between two chemicals is governing their similar hazardous properties.

Structure–hazard relationships are a ‘long-shot’: In between the structure of a chemical and its apical toxic effect are its material properties (e.g. electrophilicity), system-dependent properties (e.g. ROS generation), molecular interactions (e.g. receptorbinding and DNA-binding) and early cellular responses (e.g. mutagenicity). It is crucial to know when structure information is sufficient, or when additional data, possibly closer to the apical effect, are needed, but this should not undervalue the research efforts undertaken to derive such properties from information on structure, nor does it mean that structure and material property are unrelated. This has been exemplified with skin sensitising chemicals of low molecular weight, where reaction classes identified from the chemical structure may be a more-instructive property to predict the protein-binding than general molecular descriptors.15–17 The reaction class is considering only the property that is essential to initiate the molecular initiating event of skin sensitisation, i.e. protein binding, whereas general molecular descriptors can ‘dilute’ this information with molecular
features of less relevance.

Evidently, properties and effects closer to the apical toxic effects are more predictive and less uncertain. Lately, the concept of applying ‘functionality’ rather than (or in addition to) material descriptors was proposed for nanomaterials.18–20 This can be taken a step further: Rather than using the molecular structure or the ‘functionality’, read-across can be based on the early biological effects or common modes-of-actions of two (or more) substances. Actually, such a concept is typically represented by the common classification of any chemical with a pH of > 11 as corrosive, but no one would consider calling it functionality-based or mode-of-action-based read-across. The concept of biologically-based activity relationship (QBAR, i.e. referring to QSAR, the structure-based activity relationship) has been discussed and exemplified by van Ravenzwaay et al.12 The example of different toxicities of the structurally-similar isomers, 2-AAF and 4-AAF, was given above. These differences are reflected in different metabolome-patterns induced by these two compounds. Another example are fibrates with structural similarity. Most of these fibrates also show toxicological and pharmacological similarity, based on the metabolome data. Gemfibrozil, however, does have different pharmacological and toxicological effects. The differences in the target organ (e.g. the kidney) for Gemfibrozil and its pharmacological effect (cholesterol lowering) can be identified, based on the metabolome data. This example shows that structurally similar chemicals need not necessarily have the same apical effects, and in this case biological data are needed to prove toxicological similarity.

The call for good science and documentation in hazard assessment, that was made by Russell and Burch,2 is as relevant now as it was in 1959. Indeed, guidance documents and reporting templates have undergone several refinements,21–23 strategies have been published,11,24,25 and recently, the ECHA has published the Read-Across Assessment Framework (RAAF).1 The latter aims at the quality control and transparency of read-across evaluations. It provides structure, and ensures that all relevant elements are addressed and will lead to a conclusion on whether or not a read-across is scientifically acceptable.

Documentation and justification for a read-across approach, in a form that it is sufficient and immediately understandable for an independent reviewer, is both challenging and time consuming. It is a considerable cost factor, which is easily underestimated in the preparation of registration dossiers. In addition, a letter of access, granting the rights to use the experimental data on read-across substances, must be available. In cases where more than one study are needed, the costs for getting the rights to refer to all read-across studies may match, or even exceed, the cost of a new study. In a favourable situation, the data on the read-across substances are already owned by one of the registrants, or they have been published in sufficient detail in a peer-review journal. In this case, refusal of a read-across assessment upon evaluation is much less costly, as compared to the situation where registrants have paid a competing company for a letter of access to now-useless
read-across studies.

Read-across approaches rely on existing experimental data on potential read-across source substances. Both the generation of new data and their dissemination via the ECHA website continue to provide opportunities for read-across. Most importantly, IT tools facilitate the identification of analogues and the easy display of existing data. The most sophisticated tool in this regard is the OECD QSAR toolbox,26 but already, simpler search tools such as eChemPortal27 permit a quick search for potential read-across candidates.

Read-across has found its way in other modern chemical legislation, such as the new chemical legislations in Korea (K-REACH) and China. It helps in the hazard  assessment of new cosmetic products that are banned from animal testing in the EU. Read-across case studies are discussed at the OECD level,28 illustrating the current worldwide interest in this approach.

One of the many important points made by Russell and Burch in their 1959 book,2 is the inappropriateness of blindly taking mammalian studies as the ‘gold standard’ for human health hazard assessment. It needs to be remembered that this can also be applied to the read-across approach, since most of the experimental data on the similar chemicals are animal data. Read-across assessments predicting the outcome of animal studies may be perfect with regard to fulfilling regulatory requirements, but the ultimate aim remains human health hazard assessment.

Developing sound and well-justified read-across and grouping will be neither quick nor easy (hence it should not be termed ‘non-testing’), and it will often require fortification by ‘mode-of-action-tailored’ experimental data, in order to cover chemicals with similar early interactions, but at first glance not necessarily closely-related structures. Newly generated ‘omics’ and in vitro data addressing early (biological) effects, as well as already-existing REACH dossiers,29 SEURAT30 and Toxcast31 data, offer tools to improve read-across, based on properties closer to the hazard (the apical effect) beyond the traditional concept based only on QSARs. Established AOPs and the identification of molecular initiating events (MIEs) facilitate this use of read-across (and were, on the other hand, often identified from a set of experimental data from structurally-related chemicals). The combination of different experimental data and their relation to apical toxic effects may indeed offer the most powerful tools to advance the Three Rs. Considerations of relevant data in creating a read-across case are also used to build IATAs. Both require a sound scientific case, relevant data to support them, and awareness (and acceptance) of their limitations.

Consensus on what an acceptable read-across looks like, is emerging whilst it is in the process of being used. For this, we have to nourish and nurture the duckling — and we have to recognise when it is no longer an ugly duckling, but has matured and become
a beautiful swan (Figure 2).
Figure 1
Author for correspondence:
Dr Robert Landsiedel
Experimental Toxicology and Ecology
Ludwigshafen am Rhein

Dr Wera Teubner
BASF Schweiz AG
Product Safety


1 ECHA (2015). Read-Across Assessment Framework (RAAF) [ECHA-15-R-07-EN], 38pp. Helsinki, Finland: European Chemicals Agency. Available at: http://
(Accessed 11.11.15).
2 Russell, W.M.S. & Burch, R.L. (1959). The Principles of Humane Experimental Technique, 238pp. London, UK: Methuen. Available online: http://altweb.jhsph.
3 Berggren, E., Amcoff, P., Benigni, R., Blackburn, K., Carney, E., Cronin, M., Deluyker, H., Gautier, F., Judson, R.S., Kass, G.E.N., Keller, D., Knight, D., Lilienblum,
W., Mahony, C., Rusyn, I., Schultz, T., Schwarz, M., Schüürmann, G., White, A., Burton, J., Lostia, A.M., Munn, S. & Worth, A. (2015). Chemical safety assessment
using read-across: Assessing the use of novel testing methods to strengthen the evidence base for decision making. Environmental Health Perspectives, E-pub ahead of print [doi: 10.1289/ ehp.1409342].
4 Worth, A., Barroso, J., Bremer, S., Burton, J., Casati, S., Coecke, S., Corvi, R., Desprez, B., Dumont, C., Goul iarmou, V., Goumenou, M., Gräpel, R., Griesinger,
C., Halder, M., Janusch Roi, A., Kienzler, A., Madia, F., Munn, S., Nepelska, M., Paini, A., Price, A., Prieto, P., Rolaki, A., Schäffer, M., Triebe, J., Whelan, M., Wittwehr, C. & Zuang, V. (2014). Alternative Methods for Regulatory Toxicology — A State-of-the-art review.
Report EUR 26797 EN, 461pp. Ispra, Italy: European Commission, Joint Research Centre, Institute for Health and Consumer Protection. Available at: http://echa.
report_en.pdf (Accessed 11.11.15).
5 Cronin, M.T., Madden, J., Enoch, S.J., & Roberts, D.W. (2013). Chemical Toxicity Prediction: Category Formation and Read-across, 191pp. London, UK: Royal
Society of Chemistry.
6 Rovida, C., Alépée, N., Api, A.M., Basketter, D.A., Bois, F.Y., Caloni, F., Corsini, E., Daneshian, M., Eskes, C.,Ezendam, J., Fuchs, H., Hayden, P., Hegele-Hartung,
C., Hoffmann, S., Hubesch, B., Jacobs, M.N., Jaworska,  J., Kleensang, A., Kleinstreuer, N., Lalko, J., Landsiedel, R., Lebreux, F., Luechtefeld, T., Locatelli, M.,  Mehling, A., Natsch, A., Pitchford, J.W., Prater, D., Prieto, P., Schepky, A., Schüürmann, G., Smirnova, L., Toole, C., van Vliet, E., Weisensee, D. & Hartung, T. (2015). Integrated Testing Strategies (ITS) for safety assessment. ALTEX 32, 25–40.
7 Willett, C., Caverly Rae, J., Goyak, K.O., Minsavage, G., Westmoreland, C., Andersen, M., Avigan, M., Duché, D., Harris, G., Hartung, T., Jaeschke, H., Kleensang, A.,
Landesmann, B., Martos, S., Matevia, M., Toole, C., Rowan, A., Schultz, T., Seed, J., Senior, J., Shah, I., Subramanian, K., Vinken, M. & Watkins, P. (2013). Building shared experience to advance practical application of pathway-based toxicology: Liver toxicity mode-of-action. ALTEX 31, 500–519.
8 Anon. (2007). Regulation (EC) No 1907/2006 of the European Parliament and of the Council of 18 December 2006 concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH), establishing a European Chemicals Agency,
amending Directive 1999/45/EC and repealing  Council Regulation (EEC) No 793/93 and Commission Regulation (EC) No 1488/94 as well as Council Directive 76/769/EEC and Commission Directives 91/155/EEC, 93/67/EEC, 93/105/EC and 2000/21/EC.
Official Journal of the European Union L136, 29.05.2007, 3–280.
9 ECHA (2014). The Use of Alternatives to Testing on Animals for the REACH Regulation: Second Report under Article 117(3) of the REACH Regulation [ECHA-
14-A-07-EN], 131pp. Helsinki, Finland: European Chemicals Agency. Available at: http://echa.europa. eu/documents/10162/13639/alternatives_test_ animals_2014_en.pdf (Accessed 11.11.15).
10 Ball, N., Bartels, M., Budinsky, R., Klapacz, J., Hays, S., Kirman, C. & Patlewicz, G. (2014). The challenge of using read-across within the EU REACH regulatory framework; how much uncertainty is too much? Dipropylene glycol methyl ether acetate, an exemplary case study. Regulatory Toxicolology & Pharmacology 68, 212–221.
11 Schultz, T.W., Amcoff, P., Berggren, E., Gautier, F., Klaric, M., Knight, D.J., Mahony, C., Schwarz, M., White, A. & Cronin, M.T.D. (2015). A strategy for structuring and reporting a read-across prediction of toxicity. Regulatory Toxicology & Pharmacology 72,
12 van Ravenzwaay, B., Herold, M., Kamp, H., Kapp,  M.D., Fabian, E., Looser, R., Krennrich, G., Mellert, W., Prokoudine, A., Strauss, V., Walk, T. & Wiemer, J.
(2012). Metabolomics: A tool for early detection of toxicological effects and an opportunity for biology based grouping of chemicals — from QSAR to QBAR.
Mutation Research 746, 144–150.
13 Hagen, M., Pabel, U., Landsiedel, R., Bartsch, I., Falany, C.N., & Glatt, H. (1998). Expression of human estrogen sulfotransferase in Salmonella typhimurium:
Differences between hHST and hEST in the enantioselective activation of 1-hydroxyethylpyrene to a mutagen. Chemico-Biological Interactions 109, 249–
14 Brenna, E., Fuganti, C. & Serra, S. (2003). Enantio-selective perception of chiral odorants. Tetrahedron: Asymmetry 14, 1–42.
15 Roberts, D.W., Aptula, A.O., Patlewicz, G., & Pease, C. (2008). Chemical reactivity indices and mechanism ‐based read‐across for non‐animal based assessment
of skin sensitisation potential. Journal of Applied Toxicology 28, 443–454.
16 Teubner, W., Mehling, A., Schuster, P.X., Guth, K., Worth, A., Burton, J., van Ravenzwaay, B. & Landsiedel, R. (2013). Computer models versus reality:
How well do in silico models currently predict the sensitization potential of a substance. Regulatory Toxicology & Pharmacology 67, 468–485.
17 Urbisch, D., Mehling, A., Guth, K., Ramirez, T., Honarvar, N., Kolle, S., Landsiedel, R., Jaworska, J., Kern, P.S., Gerberick, F., Natsch, A., Emter, R., Ashikaga, T., Miyazawa, M. & Sakaguchi, H. (2015). Assessing skin sensitization hazard in mice and men
using non-animal test methods. Regulatory Toxicology & Pharmacology 71, 337–351.
18 Hendren, C.O., Lowry, G.V., Unrine, J.M., & Wiesner, M.R. (2015). A functional assay-based strategy for nanomaterial risk forecasting. Science of the Total Environment 536, 1029–1037.
19 Arts, J.H., Hadi, M., Irfan, M.A., Keene, A.M., Kreiling, R., Lyon, D., Maier, M., Michel, K., Petry, T., Sauer, U.G., Warheit, D., Wiench, K., Wohlleben, W. & Landsiedel, R. (2015). A decision-making framework for the grouping and testing of nanomaterials
(DF4nanoGrouping). Regulatory Toxicology & Pharmacology 71, Suppl. 2, S1–S27.
20 Oomen, A.G., Bos, P.M., Fernandes, T.F., Hund-Rinke, K., Boraschi, D., Byrne, H.J., Aschberger, K., Gott ardo, S., von der Kammer, F., Kühnel, D., Hristozov, D., Marcomini,
A., Migliore, L., Scott-Fordsmand, J., Wick, P. & Landsiedel, R. (2014). Concern-driven integrated approaches to nanomaterial testing and assessment — Report of the
NanoSafety Cluster Working Group 10. Nanotoxicology 8, 334–348.
21 ECHA (2008). Guidance on Information Requirements and Chemical Safety Assessment, Chapter R.6: QSARs and grouping of chemicals, 134pp. Helsinki, Finland:
European Chemicals Agency. Available at: https://
requirements_r6_en.pdf (Accessed 12.11.15).
22 ECHA (2012). Practical Guide 6 — How to report readacross and categories [ECHA-10-B-11.1-EN], 32pp. Helsinki, Finland: European Chemicals Agency. Available
at: 10162/13655/pg_
report_readacross_en.pdf (Accessed 12.11.15).
23 ECHA (2013). Grouping of Substances and Read-across Approach — An illustrative example [ECHA-13-R-02-EN], 11pp. Helsinki, Finland: European Chemicals Agency.
Available at:
13628/read_across_introductory_note_en.pdf (Accessed 12.11.15).
24 Patlewicz, G., Ball, N., Becker, R.A., Booth, E.D., Cronin, M.T., Kroese, D., Steup, D., van Ravenzwaay, B. & Hartung, T. (2014). Food for Thought: Read-across approaches — misconceptions, promises and challenges ahead. ALTEX 31, 387–396.
25 Patlewicz, G., Ball, N., Boogaard, P.J., Becker, R.A. & Hubesch, B. (2015). Building scientific confidence in the development and evaluation of read-across. Regulatory
Toxicology & Pharmacology 72, 117–133.
26 OECD (undated). The OECD QSAR toolbox. Paris, France: Organisation for Economic Co-operation & Development. Available at:
theoecdqsartoolbox.htm (Accessed 12.11.15).
27 OECD (undated). The Global Portal to Information on Chemical Substances (eChemPortal). Paris, France: Organisation for Economic Co-operation & Development.   Available at:
index?pageID=0&request_locale=en (Accessed 12.11. 15).
28 OECD (2014). Series on Testing and Assessment No. 194: Guidance on Grouping of Chemicals, 2nd edition, ENV/JM/MONO(2014)4, 141pp. Paris, France: Organisation
for Economic Co-operation & Development.
29 ECHA (undated). Registered substances. Helsinki, Finland: European Chemicals Agency. Available at: http://echa. (Accessed 12.11.15).
30 SEURAT (undated). Welcome to the SEURAT-1 website. Available at: (Accessed 11.11.15).
31 EPA (2015). Toxicity Forecasting: Advancing the Next Generation of Chemical Evaluation. Durham, NC, USA: US EPA Research. Available at:
chemical-research/toxicity-forecasting (Accessed 12.11.

Download the pdf here

The ‘Genomic Revolution’ and its Impact on Medical Research

Rehma Chandaria

The sequencing of the human genome held great promise for better
disease treatment and a concomitant reduction in animal experiments.
Neither promise has been adequately fulfilled.

In 1953, four scientists — James Watson, Francis Crick, Maurice Wilkins and Rosalind Franklin — uncovered the double helix structure of DNA, the molecule that carries genetic information in all living beings. In the 60 years since then, the ways in which we study and manipulate genes has changed considerably. In 2001, the human genome was sequenced, providing hope that common and serious conditions, such as Alzheimer’s disease and heart disease, could be predicted before they manifested clinically, and that personalised treatments for many other human conditions could be found. However, a decade on from this historic scientific achievement,  questions were raised about the lack of real clinical progress made from this vast genetic knowledge.1,2 In part, this is due to unrealistic expectations and the difficulties involved in translating discoveries made in the laboratory to clinical products that can benefit patients.

The sequencing of the human genome and better technologies for studying human genetics came with the expectation that we would no longer have to rely on animals to learn about human diseases and to find treatments. However, in reality, there has been a sharp increase in the use of genetically-modified animals, which has been responsible for the overall increase in animal procedures undertaken in the UK3 and elsewhere, in recent years.

So what actually has been achieved as a result of the ‘genomic revolution’, and what future progress can we realistically expect to see? Furthermore, is the use of animals still justified in this genomics era?

Some examples of successes arising from the genomic revolution

New research technologies

Since 1953, and particularly accelerated after the launch of the Human Genome Project in 1990, there have been significant advances in the research techniques used to study genes. Some particular highlights include the development of the Polymerase Chain Reaction (PCR) in 1985, and progress made in sequencing methods which means that the precise order of bases in DNA can now be obtained quickly and efficiently. PCR is still routinely used by researchers across the globe, to amplify a small amount of DNA to  generate millions of copies of a particular gene.

Understanding and treating genetically inherited diseases

The individual genes responsible for causing many inheritable diseases have been identified as a result of the genomic revolution. Consequently, gene therapy is now a possibility, where a faulty gene is corrected by inactivation or replacement with a functional version of the gene. In 2012,  Glybera became the first gene therapy treatment to be approved  in Europe.4 The treatment involves the delivery of an intact copy of the faulty gene in the rare inherited genetic disorder, lipoprotein lipase deficiency. Many other gene therapy treatments are in human clinical trials, although it is still an experimental
technique with several hurdles to be overcome before it can be considered as a widespread treatment option.


Cancer usually occurs as a result of mutations in the genetic sequence,5 and therefore diagnosis and personalised treatment of cancer is one of the primary successes of the genomic revolution. The genome sequencing of different cancers has revealed common mutations and patterns which determine how cancers develop.6 Mutations in genes such as BRCA1 have been identified as increasing the risk of developing breast cancer, meaning that preventative action, such as surgery to remove the breasts, can be offered to carriers of these mutations. Treatments have been developed against cancers with particular genetic profiles. For example, trastuzumab (Herceptin) is a drug mainly used to treat breast cancer, but it is only effective against cancers which overexpress the HER2 protein.7 Another example of this selective treatment is the drug gefitinib, which is only active against the 10–15% of lung cancers that carry EGFR mutations.8

Why have more treatments not been developed since the human genome was sequenced?

Of the 14,000 genes sequenced to date, only around 3800 are associated with a genetic condition that arises when the gene is faulty.9 The most common human diseases are not caused by a single gene defect, but are associated with multiple genes, each one having a small contributing effect. Other important contributing factors include influences such as diet, gender, ethnicity, age and environment, and epigenetics. Epigenetics refers to changes to the DNA molecules and the protein with which the DNA is associated to form chromatin. These epigenetic changes affect whether particular genes are switched on or off. Although every cell in the body of an organism has identical DNA, due to epigenetic modifications the characteristics of a liver cell, for example, are very different to those of a nerve cell. In addition, DNA interacts with other molecules in cells, such as RNA, signalling molecules and receptors. These complicating factors mean that someone can appear to have the genetic features of a disease such as diabetes mellitus, but not express it clinically. Therefore the idea of predicting and preventing diseases based on a person’s genetic make-up continues to be much further away than initially was hoped.

Genetically-altered animals

Since the human genome was sequenced, there has been a large increase in the use of genetically altered animals, and in particular mice, with the aim of learning about the roles of individual genes in whole animals. A genetically-altered animal has been bred or engineered to mutate, remove or insert a gene of interest. However, as the complexities already described above suggest, it is not always possible to find out the role of a gene in humans just by altering it in a mouse. Interactions with the environment and other components in a cell mean that the effects can vary greatly between species.10 In  fact, there are even examples of gene knockouts having different effects in different strains of mice.11 Therefore, we have to question whether data from genetically-altered mice can really provide insights into complex genetic interactions in humans. This is particularly relevant because of the inefficient process by which the desired genetic alterations are achieved, which results in large numbers of animals being bred but not used in experiments. There are several steps in the breeding procedure at which surplus animals are produced. The gene of interest (transgene) is constructed by using recombinant DNA technology, and then inserted into embryos which are implanted into female mice with the hope that they become pregnant and give birth to mice carrying the transgene. However, only 25% of the animals will contain the transgene, and even fewer will prove to be satisfactory for further study.12 Furthermore, extensive cross-breeding and back-crossing is required to produce animals with a homozygous genetic background in which both copies of the gene are correctly modified. The majority of the animals produced in these steps will not contain the correct genotype, are therefore considered surplus to requirements and are usually killed.13 Data from The Netherlands indicate that the number of animals that are bred but not used is almost equal to the number of animals used in experiments.14 This number of surplus animals continues to increase, due to the ongoing rise in the use of genetically-altered animals.

There are different levels of understanding human biology: the whole body, which has historically been studied by using non-human animals; tissue and organs; the cellular level; and the molecular level — which is where genomics is focused. To fully grasp how diseases develop, and to find new treatments, we need to understand the processes that happen at every level, and not just molecular processes. Despite the great successes of the genomic revolution, it is important to realise that work on genes is not the be all and end-all of biomedical research. There are complex systems and processes that occur in the body which are not related to genetics. Examples of this at the cellular and tissue levels are post-translational modifications, which occur to proteins after  they have been produced, and interactions with the extracellular matrix that surrounds cells, providing mechanical and biochemical cues that influence how cells behave. With this in mind, it is crucial to maintain efforts in all areas and at all levels of biology, rather than looking at genes and genomics in isolation. It is also vital that this is done by using scientifically- valid approaches, which provide results that are relevant to humans. This will not be achieved by using animals, and especially not genetically-modified animals. Very exciting progress has been made by biologists who have successfully used mathematical approaches to integrate data provided by in vitro techniques and thereby usefully simulate and predict physiological responses in vivo.15 This suggests that a more-complete understanding of human disease can be obtained by using human tissues and by investing in research across different levels.


From improvements in laboratory research technologies, to increased knowledge of genetic contributions to disease and genetic therapies, much has been achieved as the result of the genomic revolution. However, the resulting clinical outcome seen by patients has been much smaller than was anticipated, due to difficulties in translating research into actual treatments. This is due to the complex nature of how genes interact with other factors contributing to disease, as well as an over-reliance on animal models that are inherently different to the humans we are trying to treat. However, if the scientific community is able to turn its focus to human-based research, then the future prospects for new drugs and therapies to improve human health will be much greater.

Rehma Chandaria
Russell & Burch House
96–98 North Sherwood Street
Nottingham NG1 4EE


1 Evans, J.P., Meslin, E.M., Marteau, T.M. & Caulfield, T. (2011). Genomics. Deflating the genomic bubble. Science, New York 331, 861–862.
2 Marshall, E. (2011). Human genome 10th anniversary. Waiting for the revolution. Science, New York 331, 526–529.
3 Hudson-Shore, M. (2014). Statistics of Scientific Procedures on Living Animals 2013: Experimentation continues to rise — the reliance on genetically altered animals must be addressed. ATLA 42, 261–266.
4 Ylä-Herttuala, S. (2012). Endgame: Glybera finally recommended for approval as the first gene therapy drug in the European Union. Molecular Therapy 20, 1831–1832.
5 Loeb, K.R. (2000). Significance of multiple mutations in cancer. Carcinogenesis 21, 379–385.
6 Feng, H., Wang, X., Zhang, Z., Tang, C., Ye, H., Jones, L., Lou, F., Zhang, D., Jiang, S., Sun, H., Dong, H., Zhang, G., Liu, Z., Dong, Z., Guo, B., Yan, H., Yan, C., Wang, L., Su, Z., Li, Y., Nandakumar, V., Huang, X.F., Chen, S.Y. & Liu, D. (2015). Identification of genetic mutations in human lung cancer by targeted sequencing. Cancer Informatics 14, 83–93.
7 Yu, D. & Hung, M.C. (2000). Overexpression of ErbB2  in cancer and ErbB2-targeting strategies. Oncogene 19, 6115–6121.
8 Paez, J.G., Jänne, P.A., Lee, J.C., Tracy, S., Greulich, H., Gabriel, S., Herman, P., Kaye, F.J., Lindeman, N., Boggon, T.J., Naoki, K., Sasaki, H., Fujii, Y., Eck, M.J., Sellers, W.R., Johnson, B.E. & Meyerson, M. (2004). EGFR mutations in lung cancer: Correlation with clinical
response to gefitinib therapy. Science, New York 304, 1497–1500.
9 Barlow-Stewart, K. (2012). The human genetic code — the human genome project and beyond. Fact Sheet 24, 6pp. St Leonards, NSW, Australia: Centre for Genetics Education. Available at: http:// www. and Resources/
Genetics-Fact-Sheets/TheHumanGeneticCodeThe HumanGenomeCodeandBeyondFS24 (Accessed 30.09. 15).
10 Van Zutphen, L.F. (2000). Is there a need for animal models of human genetic disorders in the postgenome era? Comparative Medicine 50, 10–11.
11 Pearson, H. (2002). Surviving a knockout blow. Nature, London 415, 8–9.
12 Smith, K.R. (2002). Animal genetic manipulation — a utilitarian response. Bioethics 16, 55–71.
13 Combes, R.D. & Balls, M. (2014). Every silver lining has a cloud: The scientific and animal welfare issues surrounding a new approach to the production of transgenic animals. ATLA 42, 137–145.
14 Hendriksen, C. & Spielmann, H. (2014). New techniques for producing transgenic animals — a mixed blessing from both the scientific and animal welfare perspectives. ATLA 42, 93–94.
15 Butcher, E.C., Berg, E.L. & Kunkel, E.J. (2004). Systems biology in drug discovery. Nature Biotechnology 22, 1253–1259.

The Use of 3-D Models as Alternatives to Animal Testing

Hajime Kojima

A number of three-dimensional in vitro models are now available,
but significant further developments are needed before their routine
and widespread use as alternatives to animal testing will be possible

Download a pdf of this article

The development and validation of new ex vivo and in vitro test methods are urgently needed, in order to expand the use of alternatives to animal testing worldwide.  A number of such tests are already used for screening in a wide range of pharmaceutical developments, as well as in toxicological testing for regulatory purposes. These in vitro models are not commonly used, however, except to evaluate local toxic and genotoxic effects. Other toxicological fields currently utilise fish and other animals for testing, rather than in vitro or other non-animal alternatives.

I personally am hoping for the development of new ex vivo and in vitro test methods, because they are correlated with the successful development and application of regenerative medicine and tissue engineering. One important element of this research that has made significant progress is the development of novel cell types, such as cell lines, primary cultured cells, embryonic stem (ES) cells, induced pluripotent stem (iPS) cells, and mesenchymal stem cells (MSCs). There are, however, various limitations inherent in the use of cultured monolayer cells, which is why much work is currently under way in the development of three-dimensional (3-D) cell culture models. The 3-D models are superior to monolayer culture models in promoting higher levels of cell differentiation and tissue organisation, and being more appropriate because of the flexibility of the ECM (extracellular matrix) gels used, which can accommodate shape changes and intracellular connections. Rigid monolayer culture substrates are not capable of this, which is why they are not suitable for properly assessing the modes of action of medicines, toxicants and other substances.

Another important element is the development of new biomaterials for use as scaffolds for effecting proper intercellular connections. These take the form of collagen gels, spheroids and fibres, and they are fundamental for good 3-D models, which not only rely on the cells, but also on the use of the proper biomaterials. Also, at present it is difficult, if not impossible, to effect the adequate exposure of monolayer cells to substances that are not readily soluble in culture medium. Many researchers expect that 3-D models will provide a solution to such issues.

In this report, I would like to outline the current status of this research, together with both the limitations and the future potential that 3-D models represent for the development of non-animal test methods.

Recent trends in 3-D models


As early as 1970, Thomas et al. reported on the modelling of organs by using animal cells.1 Since then, many researchers have attempted to culture the liver, kidney, heart, blood vessels and various other organs, by using animal or human cells.2 Most of these
models were surrogates for external organs — including human dermis, epidermis, full-thickness and pigmented epidermis models — and a number of them are now commercially-available worldwide3, 4 for use in safety assessment and efficacy testing. These models are useful both for dermal research and for the safety assessment of skin corrosion, skin irritation and dermal absorption. The human pigmented epidermis model is used extensively in the cosmetics industry, to evaluate the whitening efficacy of new cosmetic ingredients.

Other models include the human ocular or corneal epithelium, oral epithelium, conjunctival epithelium, gingival epithelium, vaginal epithelium, bladder epithelium, intestinal epithelium, colon epithelium, alveolar epithelium,  vasculogenesis/angiogenesis5 and cardiovascular models,5 several of which are also commercially available,3,4 and are used worldwide in research and for toxicological safety assessments. The alveolar epithelium model,6 in particular, is used to assess the effects of nanoparticles, which increasingly appear in industrial products and are considered a potential cause of respiratory toxicity in humans.

There is also a significant amount of research on 3-D models of hepatocytes, based on biomaterials such as collagen gels, spheroids and fibres. Primary hepatocytes or cell lines derived from the liver are useful for studying long-term culture effects, the maintenance of functional structure, and the functional expression of the human liver. Similar liver models from a variety of animal species are being considered for use in pharmaceutical screening.

The regulatory use of 3-D models

The current Organisation for Economic Co-operation and Development (OECD) Test Guidelines (TGs) address human health hazard endpoints for skin corrosion, skin irritation, and eye irritation following exposure to a test chemical. These TGs describe in vitro procedures for identifying chemicals (substances and mixtures) not requiring classification and labelling for local toxicological damage, in accordance with the UN Globally Harmonised System of Classification and Labelling of Chemicals (GHS):7
— TG428: Skin Absorption: In Vitro Method8 This TG describes an in vitro procedure that has been designed to provide information on absorption of a test substance, ideally radio-labelled, that has been applied to the surface of a skin sample separating the donor chamber and receptor chamber of a diffusion cell. Static and flow-through diffusion cells are both acceptable for use in this assay. Skin from human or animal sources can be used. Although viable skin is preferred, non-viable skin can also be used. The absorption of a test substance during a given time period (normally 24 hours) is measured by analysis of the receptor fluid and the distribution of the test chemical in the test system; the absorption profile over time should be presented.
— TG430: In Vitro Skin Corrosion: Transcutaneous Electrical Resistance Test Method (TER)9 This TG describes an in vitro procedure that is useful for identifying non-corrosive and corrosive substances and mixtures, based on the rat skin transcutaneous electrical resistance (TER) test method. The test chemical is applied to three skin discs for a duration not exceeding 24 hours. Corrosive substances are identified by their ability to produce a loss of normal stratum corneum integrity and barrier function, which is measured as a reduction in the TER below a threshold level (5kΩ for rats). A dye-binding step incorporated into the test procedure permits the determination of whether or not increases in ionic permeability are due to physical destruction of the stratum corneum.
— TG431: In Vitro Skin Corrosion: Reconstructed Human Epidermis (RhE) Test Method10 This TG describes an in vitro procedure that is useful for identifying non-corrosive and corrosive substances and mixtures, based on a 3-D human skin model which reliably reproduces the histological, morphological, biochemical, and physiological properties of the upper layers of human skin, including a functional stratum corneum. The procedure with reconstituted human epidermis is based on the principle that corrosive chemicals are able to penetrate the stratum corneum by diffusion or erosion, and are cytotoxic to the underlying cell layers. Cell viability is measured by enzymatic conversion of the vital dye MTT (3-[4,5-dimethylthiazol- 2-yl]-2,5-diphenyltetrazolium bromide; yellow tetrazole) into a blue formazan salt that is quantitatively measured after extraction from the tissues (the MTT assay). Corrosive substances are identified by their capacity to reduce cell viability below the defined threshold.
— TG439: In Vitro Skin Irritation — Reconstructed Human Epidermis Test Method11 This TG describes an in vitro procedure that is useful for hazard identification of irritant chemicals (substances and mixtures) in accordance with GHS Category 2. It is based on reconstructed human epidermis (RhE), which in its overall design closely mimics the biochemical and physiological properties of the upper parts of the human skin. Cell viability  is measured by using the MTT assay. Irritant test chemicals are identified by their ability to decrease cell viability below defined threshold levels (below or equal to 50% for GHS Category 2). There are four validated test methods that conform to this TG. The use of this model in phototoxicity testing is described in the ICH (International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use) Test Guideline S10.12
— TG492: Reconstructed Human Cornea-like Epithelium (RhCE) Test Method for Identifying
Chemicals Not Requiring Classification and Labelling for Eye Irritation or Serious Eye
Damage13 This TG describes a test method for identifying chemicals that do not require classification and labelling for eye irritation or serious eye damage, by using a reconstructed human cornea-like epithelium (RhCE). This tissue construct closely mimics the histological, morphological, biochemical and physiological properties of the human corneal epithelium. The purpose of this TG is to describe the procedures used to evaluate the eye hazard potential of a test chemical, based on its ability to induce cytotoxicity in the RhCE tissue construct, as measured by using the MTT assay.

Future potential and limitations of the 3-D models

Future potential

A range of TGs describing test methods that use epidermal and/or ocular models are already available worldwide for regulatory use. The quality of the procedures that use these models is maintained by the suppliers. TG428 includes the use of an ex vivo skin model for assessing the effects of exposure to chemicals. In the future, in vitro full-thickness skin, intestine and alveolar models are expected to be used for assessing the effects of exposure to chemicals. It is absolutely necessary for these models to evaluate absorption at the threshold of the  physiologically based toxicokinetic (PBTK) model. On the other hand, I expect new developments for the hepatocyte model. Since 1997, the European Medicines Agency (EMA) and US Food and Drug Administration (FDA) Guidelines14,15 have required a CYP (cytochrome P450) induction assessment for new pharmaceuticals. However, human CYP induction for the safety assessment of a broad spectrum of test chemicals (e.g. cosmetics, food additives, pesticides, mixtures) is currently not systematically addressed by any OECD TG. Despite this shortcoming, the induction of CYP enzymes in monolayer hepatocytes by drugs, and the potential of 3-D models for use in this type of study, are receiving attention from researchers.

Furthermore, ‘human-on-a-chip’ and ‘organ-on-a-chip’ research focuses on in vitro human organ constructs for the heart, liver, lung and the circulatory system in communication with each other. The goal is to assess effectiveness and/or toxicity of drugs in a way that is relevant to humans and their ability to  process these pharmaceuticals. The 3-D culture models fail to mimic the cellular properties of organs in many aspects, including cell-to-cell interfaces or the complete organ as a whole. The application of microfluidics in organ-on-a-chip methodologies provides
for the efficient transport and distribution of  nutrients and other soluble items throughout the viable 3-D tissue constructs. Organs-on-chips are referred to as the ‘next wave’ of 3-D cell culture models, that mimic the whole living biological activities of organs, and their dynamic mechanical properties and biochemical functions.


Unfortunately, the current models need significant further developments, and most of them are constructed with only one cell type. Therefore, their construction and functions are not comparable to ex vivo models. I hope for further advances in these areas, particularly because 3-D epithelium models have advanced very little over the past decade. I expect the development of 3-D models of a wide variety of cell types to be achieved, and that a model constructed with differentiated cells (including different types of stem cells) will be produced in the near future — for example, a full-thickness skin model that includes melanocytes, Langerhans cells and hair follicle cells derived from stem cells. In addition, the toxicological biomarker for all of the current 3-D models, and the one that is accepted in the OECD TGs, is cytotoxicity. Actually, cytotoxicity is one biomarker, but I do not consider this to be a specific biomarker based on mode of action. Like specific CYP enzymes, specific toxicological biomarkers for each developed organ should be used. The economic viability of developing a wide variety of small-scale 3-D models remains precarious. A production platform that enhances efficiency is needed. The long lead-time required to prepare 3-D models is another factor that drives up costs. Further study of 3-D modelling by using cells differentiated from ES or iPS cells is impossible without the development of quality control criteria for the system used to differentiate the cells. It is difficult to coordinate the longterm maintenance of 3-D models with combinations of cells, and it will be necessary to co-culture with organ-derived substances and reconstructed blood vessels, in order to promote the development of humans-on-chips or organs-on-chips.

Dr Hajime Kojima
National Institute of Health Sciences
1-18-1 Kamiyoga
Tokyo 158-8501


1 Thomas, J.A. (1970). Organ Culture, 512pp. New York, NY, USA: Academic Press.
2 Antoni, D., Burckel, H., Josset, E. & Noel, G. (2015).Three-dimensional cell culture: A breakthrough in vivo. International Journal of Molecular Sciences 16, 5517–5527.
3 MatTek (2015). In Vitro Tissue Models. Ashland, MA, USA: MatTek Corporation. Available at: (Accessed 30.07.15).
4 Xenometrix (2015). Homepage. Allschwil, Switzerland: Xenometrix AG. Available at: (Accessed 30.07.15).
5 Heinonen, T. (2015). Better science with human cellbased organ and tissue models. ATLA 43, 29–38.
6 Jamin, A., Sr (2015). Predicting respiratory toxicity using a human 3D airway (EpiAirway™) model combined with multiple parametric analysis. Applied In Vitro Toxicology 1, 55–65.
7 Anon. (2015). Globally Harmonised System of Classif ication and Labelling of Chemicals. [In Japanese.] Tokyo, Japan: Ministry of Health, Labour & Welfare. Available at: (Accessed 30.07.15).
8 OECD (2004). Test Guideline No. 428: Skin Absorption: In Vitro Method, 8pp. Paris, France: Organisation for Economic Co-operation and Development. Available at:
health-effects_20745788 (Accessed 30.07.15).
9 OECD (2015). Test Guideline No. 430: In Vitro Skin Corrosion: Transcutaneous Electrical Resistance Test Method (TER), 20pp. Paris, France: Organisation for Economic Co-operation and Development. Available at: 739-en;jsessionid=4dc4tau8sk8k1.x-oecd-live-03 (Accessed 27.08.15).
10 OECD (2015). Test Guideline No. 431: In Vitro Skin Corrosion: Reconstructed Human Epidermis (Rhe) Test Method, 33pp. Paris, France: Organisation for Economic Co-operation and Development. Available at:; jsessionid=4dc4tau8sk8k1.x-oecd-live-03 (Accessed 27.08.15).
11 OECD (2015). Test Guideline No. 439: In Vitro Skin Irritation: Reconstructed Human Epidermis Test Method, 21pp. Paris, France: Organisation for Economic Co-operation and Development. Available at:
439-in-vitro-skin-irritation-reconstructed-humanepidermis- test-method_9789264242845-en;jsessionid =4dc4tau8sk8k1.x-oecd-live-03 (Accessed 27.08.15).
12 US FDA (2014). S10 Photosafety Evaluation of Pharmaceuticals: Guidance for Industry, 21pp. Silver Spring, MD, USA: US Department of Health and Human Services, Food and Drug Administration. Available at: complianceregulatoryinformation/guidances/ucm337572.pdf#search=’ICH+S10 (Accessed 30.07.15).
13 OECD (2015). Test Guideline No. 492: Reconstructed Human Cornea-like Epithelium (RhCE) Test Method for Identifying Chemicals Not Requiring Classification
and Labelling for Eye Irritation or Serious Eye Damage, 25pp. Paris, France: Organisation for Economic Co-operation and Development. Available at:;jsessionid =4dc4tau8sk8k1.x-oecd-live-03 (Accessed 27.08.15).
14 US FDA (2012). Guidance for Industry Drug Interaction Studies — Study Design, Data Analysis, Implications for Dosing, and Labeling Recommendations: Draft Guidance, 79pp. Silver Spring, MD, USA: US Department of Health and Human Services, Food and Drug Administration. Available at: downloads/drugs/guidancecomplianceregulatory
information/guidances/ucm292362.pdf (Accessed 30. 07.15).
15 EMA (2012). Guideline on the Investigation of Drug Interactions, 59pp. London, UK: European Medicines Agency.

Laboratory Animal Science and the Use of Scientific Knowledge

PiLAS Staff Writer

The World Conference on Science for the Twenty-first Century: A New Commitment, took place on 26 June to 1 July 1999 in Budapest, Hungary, under the auspices of the United Nations Educational, Scientific and Cultural Organisation (UNESCO) and the International Council for Science (ICSU). The participants produced a very interesting 7-page, 46-paragraph Declaration on Science and the Use of Scientific Knowledge, which is well worth reading.1 There were no specific references to laboratory animal experimentation, but a number of the general points made have implications in terms of this particular form of scientific activity, and deserve to be given careful consideration. For example:

– Preamble para 4: Today, whilst unprecedented advances in the sciences are foreseen, there is a need for a vigorous and informed democratic debate on the production and use of scientific knowledge. The scientific community and decision-makers should seek the strengthening of public trust and support for science through such a debate. Greater interdisciplinary efforts, involving both natural and social sciences, are a prerequisite for dealing with ethical, social, cultural, environmental, gender, economic and health issues. Enhancing the role of science for a more equitable, prosperous and sustainable world requires the longterm commitment of all stakeholders, public and private, through greater investment, the appropriate review of investment priorities, and the sharing of scientific knowledge.

– Consideration para 21: That scientists with other major actors have a special responsibility for seeking to avert applications of science which are ethically wrong or have an adverse impact.

– Consideration para 22: The need to practise and apply the sciences in line with appropriate ethical requirements developed on the basis of an enhanced public debate.

– Consideration 23: That the pursuit of science and the use of scientific knowledge should respect and maintain life in all its diversity, as well as the life-support systems of our planet.

– Proclamation 31: The essence of scientific thinking is the ability to examine problems from different perspectives and seek explanations of natural and social phenomena, constantly submitted to criticalanalysis. Science thus relies on critical and free thinking, which is essential in a democratic world.

– Proclamation 40: A free flow of information on all possible uses and consequences of new discoveries and newly developed technologies should be secured, so that ethical issues can be debated in an appropriate way.

– Proclamation 41: All scientists should commit themselves to high ethical standards, and a code of ethics based on relevant norms enshrined in international human rights instruments should be established for scientific professions. The social responsibility of scientists requires that they maintain high standards of scientific integrity and quality control, share their knowledge, communicate with the public and educate the younger generation.

The challenge to all stakeholders in laboratory animal science — be they scientists, research organisations, industries, industry associations, funding bodies, medical research charities, patient associations, politicians, governments, animal welfare activists or antivivisectionists — is to strive to work, not as isolated parties, but together, to live up to the expectations of those who met in Budapest in 1999. The question that all these stakeholders must not be allowed to avoid is this: In all honesty, is this challenge being met?

1 Anon. (1999). World Conference on Science Declaration on Science and the Use of Scientific Knowledge, 7pp. Paris, France: UNESCO. Available at:

Download a PDF of this article: CLICK HERE

Beyond the Three Rs

For a long while after the Three Rs were first proposed by Russell and Burch, anti-vivisectionists rejected the concept, on the grounds that experiments on living vertebrates which cause them pain, suffering, distress or lasting harm, were ethically unacceptable and scientifically unnecessary, so there was no point in reducing, refining or replacing them. In recent decades,
however, some organisations, such as the BUAV and PETA, have moved tentatively into the middle ground, and have made positive contributions toward the Three Rs, without comprising their fundamental beliefs.

The ultimate goal of Russell and Burch themselves was replacement, which they said, “is always a satisfactory answer”, with reduction and refinement merely being steps along the way. That was also the position of the founders of FRAME, the Fund for the Replacement of Animals in Medical Experiments, although the charity has made many contributions in support
of the other two Rs, since its foundation in 1969.

The latest issue of ATLA contains two important articles about the future of the Three Rs, as well as the latest in an important series of outstanding exposures of the insurmountable limits of laboratory animals as models of humans.

In this issue of PiLAS, Craig Redmond argues the case for replacing the Three Rs with One R (Replacement),1 but goes further in saying that only what Russell and Burch defined as absolute replacement (where “animals are not required at all at any stage”) should be considered acceptable, since relative replacement can still involve suffering, as in the use of invertebrates,
less-sentient vertebrates, or cells and tissues taken from protected animals and used in vitro or ex vivo.

Michael Balls goes further in his ATLA Comment,2 proposing that “the time has come to plan for a future where the Three Rs will have served their purpose, animal experimentation will have been consigned to history, and humane biomedical science in research, testing and education will have become the norm, for the benefit of humans and animals alike”.

Finally, the article by Jarrod Bailey in the latest ATLA issue, on monkey-based research,3 demonstrates that major molecular differences, revealed by comparative
genomics and molecular biology, underlie inter-species phenotypic disparities. The collective effects of these differences are striking, extensive and widespread, and show that the superficial similarity between human and monkey genetic sequences is of little benefit for biomedical research.
Therefore, the extrapolation of biomedical data from monkeys to humans is highly unreliable, and the use of monkeys must be considered of questionable value, particularly given the breadth and potential of alternative methods of enquiry that are currently available to scientists.

1 Redmond, C. (2014). ‘One R’ is the new ‘Three Rs’. ATLA 42, P50–P52.
2 Balls, M. (2014). Animal experimentation and alternatives: Time to say goodbye to the Three Rs and hello to humanity? ATLA 42, 327–333.
3 Bailey, J. (2014). Monkey-based research on human disease: The implications of genetic differences. ATLA 42, 287–317.

Personal Reflections on Veterinary Science Training and the Three Rs

Rosemary Elliott

Students against the thoughtless use of animals
in veterinary education need  support
to be true to their values, to use
science to uphold them, and to never
give up on advocating for the
highest standards of animal welfare

This opinion piece is something I needed to write personally, and it shocks me to notice the ten years that have passed since I enrolled as a mature-age student in veterinary science. It was a privilege to have this opportunity, and I am grateful that I can now advocate for animals in a way I never could have done before, particularly through my work with Sentient. By the time I enrolled in veterinary science, the faculty where I studied had made huge advances in the ethical use of animals in teaching, most notably through the banning of ‘terminal surgeries’. Since my time as a student, things have improved even more, and the faculty is leading the way in developing a national curriculum for animal welfare and ethics in veterinary teaching. Yet, despite this, some attitudes and practices ‘die hard’, and many of my experiences could only be described as vicarious trauma — hence the passing of so much time before I felt ready to give life to them with words.

Although the animals I speak of were used primarily for educational purposes, I do hope that those of you working in the laboratory setting with research animals, will find that my reflections will stimulate your own thinking and resonate with some of your experiences. We have much in common in our struggle to do our work and to do the best we can by the animals in our care.

Anatomy classes

Admittedly, I grew up with an idealised view of veterinary practice, starting with Dr Doolittle as a very small child, then moving on to All Creatures Great and Small and the other James Herriot books. Although 60 years had passed since the setting of those stories, with the onset of intensive farming, I went on assuming it was primarily about supportive teamwork and above all, empathy for animals.

The first wake-up call came in the form of Anatomy 1A. The laboratory, a huge room with grimacing dead greyhounds lying on metal trays, became a hothouse of anxieties, sometimes played out by the larking about that involved the throwing of body parts. The smell of formalin was so powerful that I can still conjure it up. In groups of three or four, we learned anatomy the traditional way, which some of us described as “being thrown a dead dog and a textbook”. There was minimal instruction, so we were left to hack away in the pursuit of identifying a lengthy page of anatomical features, ‘all examinable’, with the pressure to clean up before our class ended.

I don’t remember being told why they were always greyhounds, or how they were sourced. There was certainly no time for debriefing, and we learned early on that it was unacceptable to appear emotional. I soon became desensitised to this horrible scene until the 4th year, when I was initiated into surgical training on freshly-killed pound dogs of various breeds — soft and floppy and somehow, not being greyhounds, they seemed more individual, more like pets. I remember struggling with sad moments of wondering about their lives and whether they had been loved, and feeling ashamed of my tacit acceptance of greyhounds as production animals.

Now, this raises for me the question of respect for life, even after a life is over. If we are serious about honouring the dignity of animals and safeguarding their welfare, veterinary training must provide as many opportunities as possible for effective learning that reduces the number of animals used, such as initial skills training through videos and silicone simulator anatomical models. At Sentient, we also call for the veterinary profession to supply cadavers from ethical sources, rather than colluding with the widespread disposal of racing greyhounds by ‘convenience euthanasia’.

Practical animal husbandry training

Chickens and eggs
The Animal Husbandry practical classes on production animals were another shock. I expected state-of-the-art facilities for the animals, but the picnic atmosphere was killed for me by the sight of housing systems more typical of factory farming. The laying hens were all kept in battery cages, rows of little prisons with no natural lighting. Questioning this was frowned upon — we were there to show our enthusiasm and tick the boxes for the mastery of required skills. I was told these were working farms, the message to first year students from some staff members being that animal welfare is, by default, secondary to industry profit.

One of my saddest memories was being tested on my ability to remove a hen from a battery cage, restrain and examine her, then to return her to the cage, head first. She flapped with gusto and resisted going back inside. I felt I had performed an act of cruelty, giving her a tiny taste of what her body could do, but perhaps never would again. Then there was the class where chicks were deliberately infected with coccidiosis, so that we could observe the characteristic droppings. And the class on egg production, where the tutor smilingly broke fertilised eggs — to prove what, I don’t understand anymore — but leaving me with the image of an embryonic bird with a throbbing heartbeat, destined for the sink.

Compassion for birds was a disadvantage in veterinary training, where speciesism was very much alive and well. Most student complaints about animal treatment focused on mammals. In the 4th year, we were taught to euthanase chickens humanely, but were given needles of the wrong gauge, which caused their wing veins to blow. I felt murderous at the sight of students who laughingly persisted, while the ‘spent’ hens blinked helplessly as they were repeatedly traumatised by needles that failed to bring an end to their joyless lives. The demonstrator did not intervene. I learned nothing useful from this class — because after observing the other students, I decided not to even try. When I complained, the response was that they had run out of the correct-sized needles and would replace them at the next class, as if the problem had been purely a practical one.

This class could have formed a foundation for trainee veterinarians in the careful preparation and respect for birds during euthanasia. Instead, it reinforced the view of chickens as somehow less sentient, due to their status as production animals. We must consider how invoking the Three Rs would have helped here. Perhaps by the initial replacement of live birds by video footage of how to correctly perform euthanasia? But when it came to the use of live birds, which was essential, what was needed was a commitment by staff to the humane treatment of animals, and an expectation that students would demonstrate this attitude through their own behaviour. Instead, the atmosphere was cavalier and callous, and not at all conducive to any form of refinement.

Tail-biting management classes
The university’s pig farm was another indictment, with sows in sow stalls, and non-breeding pigs in dark little pens where we practised catching them with snares. This was the scene of my own most shameful memory, where, despite my opposition to unnecessary invasive procedures, I performed teeth-cutting, tail-docking and ear-notching on a piglet. I remember feeling terrified at the thought of this practical class, because of the expectation to ‘do as farmers do’ — which was ostensibly to improve the welfare of the piglets by preventing tail biting. I was unaware of any conscientious objection policy at that stage of my training. So I stalled for time, trying to look useful without actually doing anything, until the tutor presented me with my own piglet and a pair of what looked like pliers. He was clearly annoyed by my questions and my need for reassurance about avoiding the ear vein or how to create the least degree of trauma — a ridiculous question, in view of the fact that there was no analgesia.

I will always remember holding this warm, pink little being with the racing heart, who I just wanted to protect, but instead, I brutalised him. I try to cope with this memory by knowing that I worked as quickly as I could, and then rubbed his body all over to distract him from the pain, as I carried him back to his mother and made sure he found a teat to suckle on. I also try to cope by using this as the basis for my commitment from that day onwards to never again perform such an atrocity. But I will never forget standing there, praying the pain would soon end. I will never forget being inconsolable for days. I still cry at the memory of what I did. I betrayed my own values out of fear of failure and not being confident enough to stand up to pressure, aided by false reassurances that I was doing the right thing.

That night, I researched teeth cutting, and found a recent article in a veterinary journal that documented how the procedure predisposes to injury and infection in the mouth and gums. I have since visited free-range pig farms, where the ‘cannibalism’ we were indoctrinated with was not an issue. All these pigs kept their tails, and their baby teeth. I had been fed a lie, pressured into performing unnecessary and inhumane procedures on a piglet that would never be allowed on a puppy. And I was an educated adult, who had already considered these welfare issues and was not relying on this to earn a living.

My pathetic attempt at reduction, by limiting my actions to one piglet, did nothing to safeguard the welfare of that individual. Like so much of what we were expected to do in veterinary training, it simply should not have been allowed, because the procedure itself was unethical. In such cases, I believe we can be guided by the Three Rs to give priority to replacement, by showing students a video of these procedures, not least so that they know what to expect on farm visits. This can go hand-in-hand with advice about how these practices can be avoided through less-intensive husbandry practices. When invasive procedures are required, the focus should then be on initial video or simulator learning, to reduce the number of animals used, followed by refinement — i.e. teaching students to always use anaesthesia and/or analgesia to minimise stress and suffering.

Other lessons
There were multiple instances, where students at the university farm were expected to perform unnecessary invasive procedures on production animals as part of their learning experience, with routine farming practices cited as the ‘gold standard’. My question was always this: Why are we following the ways of the farmers, rather than offering something more as potential veterinarians? But most students, particularly those from rural backgrounds, accepted the status quo, citing the need for ‘real-world’ practice. It was also regarded as essential preparation for extramural farm placements, which brought further horrors that I will not elaborate on here. Common sights during my training were cattle being dehorned without analgesia, in some instances leading to maggot-infested wounds; the repeated attempts to lasso a terrified cow on a hot summer’s day, which led to her jumping a fence and almost breaking her hip; exsanguination of a sheep to prove the procedure is humane; and tail-docking and castration of lambs, who were pinned down on their backs in a ‘cradle’, all without analgesia, while students delayed the procedure by joking about who would remove which testicle.

Promotion of the Three Rs and alternatives
My reason for writing this, apart from my own need to debrief and reflect, is to think about the context that must be created, if the Three Rs are to be of use in veterinary education. Adherence to the Three Rs will only come about within a culture of empathy and respect for animals, which should also be extended to students. Offering a transparent conscientious objection policy, and reinforcing in every unit of study that there are alternatives to ethically-contentious procedures, is a crucial part of this. The way I coped was to find like-minded students, involve myself in the student-run animal welfare association, use the conscientious objection policy and formal avenues within the faculty to lodge complaints, and seek the support of staff members. And I will always be grateful to several academics, who were wonderful role models on how to uphold ethics in teaching, who had witnessed and objected to far worse in their own veterinary education, and who encouraged me and my fellow students to be true to our values, to use science to uphold them, and to never give up on advocating for the highest standards of animal welfare.


Our Guiding Principles

Veterinary surgeons should let themselves be guided by
the principle “In dubio pro animale” — roughly translated as
“When in doubt, support the animals”.

Katherine van Ekert Onay

One of my most challenging career experiences occurred during my first time at an animal-testing institute. Mice, crippled with advanced-stage arthritis, were intentionally deprived of analgesia, due to concerns about the impact of analgesics on the experimental results obtained. Neurosurgery, which requires, in the human world, performance by only the highest calibre of neurosurgeons after decades of training, was routinely performed on rats by nominally trained researchers, under non-sterile conditions. The ensuing tissue aggravation surrounding their electrode skull caps was nevertheless a mere drop in the ocean compared to what I could onlyassume was an astronomical headache induced by the intrusive electric implants. Yet, despite all of this, there was again, no analgesia.

As a veterinarian drawn to the profession, like many, out of a desire to care for animals, this incongruity unsettled me. The  researcher had a clearly genuine  tenderness for her rats; lovingly embracing them in her arms as if they were her own children. But can we really call ourselves animal carers, when we allow mechanisation and complacency to perpetuate what is, in any rational context, cruelty — wanton or otherwise? Faced with these situations, we must reach inside ourselves with honesty, and ask what we are really doing this for, where our priorities lie, and whether our responsibilities toward the research arena can be genuinely harmonious with our primary responsibility toward animals?

With concern for the welfare of animals on the rise, and recent bans on the use of animals for cosmetics testing in the EU, Brazil, Israel and India, we, as animal professionals, are provided with interesting dilemmas as to how we apply ethical principles to our own practice. We are at the front line when it comes to ethical debates regarding animals, but are correspondingly well-placed to provide leadership over how they should be treated — providing our scientific knowledge and practical experience to help society make informed decisions. Indeed, one of the key principles in taking our veterinary oaths is to uphold animal welfare. Yet I wonder whether we are doing enough to challenge the status quo in research? It is only after doing so that we can confidently and sincerely uphold our role as advocates for these species. We are guilty of hypocrisy, in endorsing legislation that permits surgically invasive procedures to be performed by ‘competent’, not ‘qualified’ technicians in the research arena, when this is normally the sole domain of a veterinarian. We apply double standards to how we accept pain,  approving projects that cause animals to experience “pain and/or distress that will not be alleviated”; and assuming that because a mouse is eating and is relatively mobile, then they must be coping just fine.

Less explicitly, we dishonour the Three Rs principles of Russell and Burch with every day that we fail to disrupt our outmoded methods of sharing knowledge, which is currently safeguarded in the rigidly inaccessible silos of research faculties and institutions. As they stand, these ‘silos’ render it impossible to prove accurately the lack of viable alternatives to animal models and the need for the development of these alternatives. At least they should give us the opportunity to alleviate our concerns that we are doing ‘the best we can’, and that animal suffering is therefore inevitable, albeit unfortunate. As well as expanding our communication with research faculties and institutions, we need a greater reverence for the dissemination of negative results. This would represent a  monumental move toward a reduction in animal wastage. Continuing this rigidity and inertia leads not only to delays in the acquisition of knowledge, but also causes me to question whether we are simply propping up the animal research industry in the interests of our continued employment and professional recognition, rather than serving the interests of animals, as we had originally intended.

But exploring the ethical implications of our personal actions can be challenging. As scientists, we find it safe and part of our natural territory to try to isolate science from any moral judgement about how we treat animals. A great example of this is found in the European Union’s Scientific Committee on Animal Health and Animal Welfare, Welfare Sub-Committee Mandate, which states that it is their goal to address “scientific and technical questions concerning the protection of animals, notably in regard to husbandry, herd management, transport, slaughter and experimentation.”1 [The italics are the current author’s.]

Although an essential component, science alone is not sufficient, given that we work with beings who have their own inherent value and requirements. It is not only our moral responsibility, as animal professionals, to ensure that we see them beyond the utility they serve, but we also owe it to ourselves to honour the empathy that drove us to work with animals in the first place.

The 2010 amendment to the AVMA Veterinarian’s Oath provides some  encouragement through its acknowledgement that animal welfare is a priority for the veterinary profession. The newly revised section reads: “Being admitted to the profession of veterinary medicine, I solemnly swear to use my scientific knowledge and skills for the benefit of society through the protection of animal health and welfare, the prevention and relief of animal suffering, the conservation of animal resources, the promotion of public health, and the advancement of medical knowledge.”2 [The italics are the current author’s.]

It was this interest in challenging the dissonance and complacency that we ourselves had experienced in our veterinary training and work, that resulted in the formation of Sentient, The Veterinary Institute for Animal Ethics. The mission of Sentient is to offer scientific reflection within an ethical framework to work for improvements in animal protection. We provide a voice, and a solid, evidence-based platform that helps veterinarians to independently play their role in addressing animal welfare needs.

Sentient also promotes an ongoing commitment to the highest standard of care for animals used in research. Our philosophy is that animals should be treated with respect, befitting their status as sentient beings, rather than as mere tools for data collection.

We oppose any research involving animals that causes unrelieved pain, suffering, distress, lasting physiological harm or behavioural disturbance, or death as an endpoint. This means a priori rejection of some research proposals, including: a) the use of animals to test non-essential products (such as cosmetics, household products, alcohol or tobacco); b) procedures causing neurological damage; c) the in vivo exposure of prey animals to predators; or d) the use of great apes in research other than non-invasive observational studies with free-living or sanctuary populations.

The German Veterinary Association for the Protection of Animals’ Code of Conduct perhaps best sums up what animal professionals must do in these difficult circumstances. When in doubt, veterinary surgeons should let themselves be guided by the principle, “In dubio pro animale” — roughly translated as, “When in doubt, support the animals”.3 As society places increasing value on animal welfare and sentience, it will be up to animal professionals to maintain our standard as safeguards of the animals’ best interests.

1 Scientific Committee on Animal Health and Animal Welfare (undated). Mandate.  Brussels, Belgium: European Commission, DG Health and Consumers.
Available at:
index_en.html (Accessed 06.06.14).
2 AVMA (2014). Veterinarian’s Oath. Schaumburg, IL, USA: American Veterinary Medical Association. Available at:
veterinarians-oath.aspx (Accessed 06.06.14).
3 TVT (1998). Codex Veterinarius. [Ethical guiding principles for veterinary behaviour with respect to the welfare and protection of animals (Edition from 1st June 1998).] Bramsche, Germany: Tierärztliche Vereinigung für Tierschutz e.V. Available at: (Accessed 06.06.14).

Animals with Human Rights

dog-18965966_l - Copy



Historically, cats and dogs have simply been people’s property, but their status may be changing, at least in the USA, where they can now inherit legacies and can have lawyers appointed to represent them. As a result of a number of decisions in US courts, they are said to be inching closer to rights previously thought to be only available to humans. This has many implications, as is explored in a newly-published book by David Grimm, entitled Citizen Canine: Our Evolving Relationship with Cats and Dogs.1

These developments are worrying the National Association for Biomedical Research (NABR), the US equivalent of the UK’s Understanding Animal Research, as reported by Grimm in an interview with Scientific American, subtitled Animals with Rights Will Be More Than a Pet Peeve for Researchers.2 NABR have set up an “animal law monitoring project”, and the fear is that, since there is no reliable legal distinction between companion animals and laboratory animals, the use of cats and dogs in laboratories may be threatened. Even worse, “similar
legal arguments [about animal rights] could be applied to chimps and monkeys — and then rats”.2 As Grimm says, “Opposition to animal testing has been rising among the public”, and viewing “animals as more than property is probably going to be enshrined in more policies and practices”.2

The truth is that this is all just a red herring. The discussion should not be about animal rights, but about human responsibilities. Even if animals had rights, they would not have the power to exercise them. Power over them is in our hands, just as some humans can have the power to deny rights to other humans. It is humans who decide how animals should, or should not, be treated.

There are many different reasons for questioning the use of animals in experiments which may cause them pain, suffering, distress or lasting harm, including scientific, ethical and logistic reasons. Two of the ethical considerations apply specifically to cats, dogs and Equidae, as companion animals, and to chimps and monkeys, which have special cognitive abilities and a capacity to suffer which are very close to those of humans. These important considerations are discussed by Hubrecht in a new book on The Welfare of Animals Used in Research.3

One argument for limiting the use of cats, dogs and Equidae, is that they have evolved in parallel with human beings over thousands of years, leading to the development of close relationships of various kinds, as our friends and companions, but also as our servants. Dogs, for example, serve us as guides, guards, detectives and farm labourers. They may not suffer more than several other laboratory animals, but our special relationship with them imposes an extra burden of responsibility on us. We owe them gratitude, rather than further exploitation.

1 Grimm, D. (2014). Citizen Canine: Our Evolving Relationship with Cats and Dogs, 352pp. New York, NY, USA: Public Affairs.
2 Fischman, J. (2014). Animals with human rights make researchers run scared. Scientific American, 17 April 2014, 2pp. Available at: http://www.scientificamerican.
com/article/animals-with-human-rights-will-be-more-than-a-pet-peeve-forresearchers/?&WT.mc_id=SA_WR_20140423 (Accessed 28.04.14).
3 Hubrecht, R.C. (2014). The Welfare of Animals Used in Research, vii + 271pp. Chichester, West Sussex, UK: Wiley Blackwell.

Locating the Middle Ground

A ‘middle ground’ exists, where the advocates and
the opponents of vivisection can usefully negotiate,
but where exactly is it?

In editorials for ATLA, and in his recent pieces for  PiLAS, Michael Balls has sometimes spoken about the  ‘middle ground’ which exists “between the extremes  and biases of the most trenchant anti-vivisection or  pro-vivisection positions”.1 This middle ground, he has argued, is where useful negotiation takes place,  as indeed he and others demonstrated in the case of the Animals (Scientific Procedures) Act 1986.2 But Professor Balls is not the only person to use this terminology, and like any busy metaphor (and one needs reminding that it is a metaphor), it requires periodical re-definition, if it is to remain helpful. I propose to offer such a re-definition here.


We might start with the ‘extremes’ — i.e. the far limits which fix the middle ground. For ease of visualisation, I shall speak of ‘left’ and ‘right’, though without any party-political meaning. On the left, then, there is abolitionism, and it must be remembered  that this is an ‘extreme’ only in so far as, morally or intellectually, there is nowhere further leftwards to go. For we are talking here about ideas, not persons, a most important but often blurred distinction. The term for the person, i.e. ‘extremist’, can be used to mean someone dissatisfied with compromise (as indeed an abolitionist is), but it is almost always used to mean someone who goes to extremes. So this word has somewhat discredited the other, which is why it is necessary to reclaim the strictly positive or geometrical sense of ‘extreme’: the sense which makes, for instance, a claim for equality in human rights extreme, because that is the utmost that can be achieved in that direction, rather than because it is an unreasonable claim in itself.

So we know where we are on the left. To clarify the right extreme, I quote the philosopher, R.G. Frey, writing in The Oxford Handbook of  Practical Ethics: “To many people, of course, there has to be a middle ground between the abolitionist and anything goes positions.”3 This is surely a misrepresentation — although since it appears thus in an authoritative survey, it is probably a widely-accepted one. The view that “anything goes” was in fact taken off the  map in 1986, if not in 1876. Put into practice, it would be illegal, at least in the UK. As an opinion with no practical counterpart, then, it has ceased to play a useful part. And that brings us back to the point that the extremes control the location of the middle ground. That was indeed where the 1986 Act was negotiated, but the Act’s crucial achievement was to pull the right extreme inwards, and to shift the middle ground leftwards accordingly.

UK institutions which use animals in experiments, often seem to overlook this change. They will express pride, for instance, in their ethical pre-assessment of experiments, in their efforts to minimise animal suffering, and in their preference for alternatives to animals. Reassurances of this kind appear in their collective ‘Declaration on Openness on Animal Research’  of 2012.4 But to claim all of that is not to make any concessions leftwards; it is simply to confirm that these institutions are obeying the law. The practices which they describe actually position them on what is now the ‘extreme’ right.

This is not logic-chopping; it affects how people think and decide about animal research and its future. Because Frey retains the now-obsolete right in his picture — the “anything goes” right — he can view the ideology which is actually a part of the present UK law as constituting the fair-minded middle ground. He accordingly places FRAME itself on the abolitionist left. And since he concludes that this “middle ground, adequately defended, constitutes … a moral justification of animal experimentation”, he effectively freezes the scene in the form in which he pictures it.3

This last point is perhaps the most important reason for keeping the territorial metaphor well watched. It must be felt and allowed to be unstable or dynamic, so that the progressive character of the 1986 Act is kept clearly in mind. The Act itself should, if properly observed, be steadily easing the middle ground leftwards. As Professor Balls has said, many scientists are themselves committed to such progressive change.2 But the annual Home Office statistics regularly record that this change is not happening — partly, no doubt, because the Act is not being carried through with sufficient determination, partly because new reasons, sound or otherwise, for using animals in research are coming in to more-than-fill the space made by any previous successful ‘replacements’.
This is where the left extreme becomes so serviceable. True, it represents a distinct cast of thought, and not just a variation on rightwards thinking, in so far as it is essentially and even exclusively ethical, as opposed to science-with-ethics. But then the ambition to satisfy some ethic or other absolutely, however difficult or even impossible that may seem (and abolition is not impossible), provides the necessary improvement-motive in every area of moral life. It certainly did so in the earlier history of vivisection in the UK, and now that the desired progress seems to have stalled, that motive, which the abolitionist left supplies, needs renewing and re-incorporating.

I say this with some personal experience in mind. Oxford University’s ethical review committee rejected co-operation with the University-based pressure group to which I belong, on the grounds that our commitment to abolition meant we could not contribute anything useful. But let’s review our metaphor. This isn’t a tug-of-war between left and right, with only the strained rope and a referee in the middle. The middle is, or ought to be, hospitable ground to all who wish to help medical science move toward what the relevant EU Directive calls its “final goal”, i.e. animal-free medical research.5 We all supposedly hope to end up at the far left. Meanwhile, let’s know where the middle ground really is, and make good use of it, whichever direction we come from!

Dr Matthew Simpson
Voice for Ethical Research at Oxford
3 Dove House Close
Oxford OX3 8BG


1 Balls, M. (2013). The Wisdom of Russell and Burch. 7. The factors governing progress. ATLA 41, P68–P69.
2 Balls, M. (2012). The conflict over animal experimentation:
Is the field of battle changing?’ ATLA 40, 189–191.
3 Frey, R.G. (2003). Animals. In The Oxford Handbook of Practical Ethics (ed. H. LaFollette), pp. 161–187. Oxford, UK: Oxford University Press.
4 Various (2012). Declaration on Openness 2012, 1pp. London, UK: Medical Research Council. Available at:
MRC008900 (Accessed 29.04.14).
5 Anon. (2010). Directive 2010/63/EU of the European Parliament and of the Council of 22 September 2010 on the protection of animals used for scientific purposes. Official Journal of the European Union L276, 20.10.2010,

Unnecessary Experiments on Dogs are Wrong

the preclinical testing of pharmaceuticals in dogs cannot
be justified on scientific or ethical grounds

beagle crop
The Times of 15 March 2014 contained a report by Tom Whipple, 1 entitled Defences come down as animal testers dare to say they’re proud. He recorded a visit to the Harlan beagle facility, near   Cambridge, where, having seen that “a room of 100 three-week-old beagles is a room of shambolic stumbling cuteness”, he turned to the “rather more orderly process” of training puppies.

“One after another”, he reported, “beagles are asked to sit, stand and offer their forelimb. Then they move to the more advanced training: offering up their jugular — so they can get used to having blood taken; sitting still with a mask on their face — so that they can receive medication; allowing clippers to be run over their neck — in case they need to be shaved. When the training is complete, they will be ready to send out for medical experimentation.”

An editorial in the same issue of The Times, entitled Animal Rights and Wrongs; the harsh truth is that vivisection is often the lesser of two evils,2 said that, given that animal experimentation “has helped deliver, along with numerous other pharmaceutical victories, a vaccine for rabies and a means of regulating diabetes, few people would be willing to sacrifice these triumphs to prolong the life of a beagle. Or a thousand beagles, come to that. In the moral calculus that has to be performed, most people would come down on the side of the experiments.”

As expected, these two pieces have led to a number of widely-different reactions. Not surprisingly, they were welcomed by Understanding Animal Research in various ways (e.g. 3) and roundly condemned by others (e.g. 4). Whether or not it is morally acceptable to use a species widely considered to be man’s best friend, and which unselfishly provides us with support in various ways, such as in guiding the blind, is a legitimate and important subject for debate. However, a no less important consideration is whether the use of thousands of dogs in medical research and drug testing is scientifically justifiable and can be considered to be an unavoidable necessity. Such justification, totally ignored by Whipple and The Times leader writer, is an inescapable requirement, before laboratory experiments on dogs become ethically acceptable and legally permissible. Despite this requirement, regulatory agencies worldwide require preclinical testing in a rodent and a non-rodent species, in attempts to ensure that new pharmaceuticals are effective and sufficiently safe for use in humans.

The non-rodent species used is usually the dog, but there is little supportive evidence of the value or necessity of such use. Recently, Bailey et al.5 evaluated an extensive data set of 2,366 drugs, for which both human and animal data were available. They concluded that the absence of toxicity in dogs provided virtually no evidence that adverse drug reactions would also  be absent in humans, so the preclinical testing of pharmaceuticals in dogs cannot be justified on scientific or ethical grounds, i.e. it cannot be considered to be necessary.

It is to be hoped that the work of Bailey et al. will lead to a re-evaluation by toxicologists, pharmacologists, pharmaceutical companies, regulators and governments, of the need for preclinical studies in dogs. It is also hoped that this re-evaluation will be combined with greater efforts to develop and validate human-oriented, non-animal test methods and strategies. In the meantime, while words such as evil are best left to leader writers, what is unnecessary cannot be right, so it must be wrong.6



1 Whipple, T. (2014). Defences come down as animal testers dare to say they’re proud. The Times, 15 March

2014. Available at:

2 Anon. (2014). Animal Rights and Wrongs. The Times, 15 March 2014. Available at:

3 Anon. (2014). Openness in animal research news. Understanding Animal Research, 19 March 2014. Available at: /03/openness-in-animal-research-news/

4 Andrews, J. (2014). Monsters. Dissident Voice, 22 March 2014. Available at:

5 Bailey, J., Thew, M. & Balls, M. (2014). An analysis of the use of dogs in predicting human toxicology and drug safety. ATLA 41, 335–350.

6 Readers of PiLAS are invited to comment on this editorial and/or on the literature referred to.