All posts by annej

Young Researchers — The Ethical Challenge

Rebecca Ram

Never before has there been a better opportunity for young researchers
to focus on replacement alternatives, not only to save animals
from unnecessary pain and suffering,
but also to pave the way for a career in cutting-edge innovation

Introduction

The Lush Prize awards and encourages individuals or organisations who have contributed to the global initiative to end animal testing, in the fields of science, public awareness, lobbying and training, and also aims to support young researchers who wish to develop a career in animal-free toxicology. The total yearly prize fund is £250,000. The ‘Young Researcher’ category of the Lush Prize welcomes nominations from early-career scientists who are keen to progress in research without animal testing. The award offers four £12,500 bursaries, to reward research and development specifically in methods for the entire replacement of animals in toxicity testing.
In this article, Lush Prize refers to testing without the use of animals by using the term ‘non-animal’ methods. They are validated, scientifically robust methods of safety testing in their own right. The use of terms such as ‘alternatives’ or ‘replacement’ methods (while useful for clarity sometimes) may suggest that animal testing is the ‘gold standard’ of safety testing, when much of the scientific industry, along with a wealth of research evidence, confirms that, aside from the suffering involved, animal tests do not reliably predict human responses. In addition, the term ‘alternatives’ is also used to describe the Three Rs methods,1 previously summarised as follows:
Refinement: to minimise suffering and distress to animals;
Reduction: to minimise the number of animals used; and Replacement: to avoid the use of living animals. Whilst reduction or refinement methods are positive steps, they are not achievements according to the ethos of the Lush Prize. We consider only the final ‘R’ (Replacement) to be a genuine alternative, as the other two Rs still involve animal use.

A brief review of previous findings

To discuss the relevance of the Young Researcher prize, a number of key animal protection organisations were contacted and interviewed during previous research for the Lush Prize. Some of these contacts are quoted and provide useful updates throughout this article. The key messages are:

— Although the acceptance and recognition of new technologies is growing at an encouraging rate, animal-free toxicology is still the ‘less travelled’ path, and any early-career researcher trying to progress in this area is likely to meet at least some resistance or challenges along the way. That said, the environment for the discussion of alternatives to animal use is expanding, so it is vital to be persistent and remain true to one’s values in pursuing career and networking opportunities.
— A very proactive attitude is needed; ethical scientists must actively seek out their opportunities, but the rewards can be hugely successful.
— Continuing to promote the anti-animal testing message to all relevant individuals in the young/early-career researcher field, as well as communicating this message at an earlier stage of education, is vital.

One of the key positive findings from interviews with previous Young Researcher prize winners is that “Young scientists don’t always have the prejudices about animal testing being the ‘best’ way of doing things”.2 An unbiased view and fresh perspective on cutting-edge science is essential, combined with raising awareness earlier in the educational system. There is also a strong link between the Young Researcher and Training prizes, as the latter awards are relevant to those involved in the education of a number of audiences, from children in early-stage schooling, through GCSE/A-level, to the undergraduate and postgraduate levels and beyond. There is considerable scope for early-career researchers working in a broad range of scientific or technical fields to get involved in non-animal methods. This is coupled with the fact that the in vitro toxicity testing market is projected to be worth $17,227 million by 2018.3 The EU cosmetic testing ban has played a key part in driving this growth.

Mainstream funding and development of Replacement methods

The importance of funding offered by the Lush Prize continues to grow. This is especially relevant to the Young Researcher Prize, as the bursaries awarded will directly fund the development of methods to replace the use of animals in ‘frontline’ research. As highlighted in previous research conducted for the Lush Prize, lack of finance is a major obstacle to the availability of non-animal methods. The ongoing reliance on animal research means that it continues to receive the vast majority of the available funding. Furthermore, those interested in non-animal research, not only have to maintain the momentum on their specific ideas and methods, but also face a need to continually look for funding or sponsorship, which ultimately impacts on the amount of time they directly spend on their research, as acknowledged by PETA in an interview with Lush Prize in 2012: “…people may have good ideas about non-animal methods, but they’re continually going to be seeking support…and funding for those”.4
To provide some figures to illustrate the above points, in the UK in 2012–2013, over £300 million of public funding was spent on projects which  “include an element of animal use”.5In contrast, a sum of just under £9 million was awarded to Three R projects broadly termed as ‘alternatives’, with the NC3Rs awarding £7 million of this total.5 The NC3Rs state that, of the funding they provide, “around 55 per cent of research awards are directed primarily at replacement, 25 per cent for reduction and 20 per cent for refinement”.6            Therefore, within this £9 million, a much lower sum was awarded to genuine non-animal (Replacement) methods, as a significant amount of ‘alternatives’ funding is donated to the other two Rs, which still involve animal use. For example, previous NC3Rs funding includes projects which develop scales for recognising facial expressions of pain in monkeys7 or facial grimace scales in rabbits.8 To add further perspective, since it was established in 2004, the NC3Rs has awarded just over £37 million in project funding. Based on the above figures, this equates to 12% of just one year’s Government funding of projects which include animal research.
Research carried out by the BUAV in 2013 revealed the stark lack of funding devoted to alternatives to animal testing across the EU Member States. Just
€18.7 million were devoted to methods relating to  the Three Rs in 2013, by only seven countries, with most Member States failing to assign any funding at all, and half of them not responding to the survey. Given that the available figures for 2011 show the total combined annual science R&D (research and development) budget for the EU to be almost €257 billion, the amount spent on alternatives is wholly inadequate, equating to just 0.007% of the total expenditure.9
These disappointing figures demonstrate the importance of independent funding for non-animal research, such as that which the Lush Prize offers. At its launch in 2012, the £50,000 total prize money for the Young Researcher Prize was allocated to five potential winners. This has now changed to award four prizes of £12,500, in order to increase the funding awarded to each individual, whilst still recognising the work of several researchers. Feedback from previous prize winners has indicated that these bursaries provide a meaningful amount, so the slight increase in funding across four awards will be of even more benefit, to go toward both research expenses and the cost of consumables.

Current and ongoing opportunities for keen young researchers

Banning animal testing will stifle innovation?

The claim that a ban on animal testing would stifle innovation was regularly made by industry as the 2013 EU cosmetics testing ban came into effect.10 Far from impeding research, the ban (both the 2009 and 2013 phases) had the opposite effect, and was the direct driver for the launch of new research into non-animal methods through large-scale, multinational projects (e.g. ReProTect11) as these two critical deadlines approached. R&D on new methods is innovation in itself, and it provides the perfect opportunity for those who genuinely want to be involved in cutting-edge, next-generation science, without causing animal suffering. The EU has led the way in progress on the development of alternative methods of testing to animals, and is considered ‘a leader in innovation’, something which should be reflected in the opportunities it offers young researchers and emerging graduate scientists.

Toxicity testing is toxicity testing, regardless of purpose

The validated and accepted non-animal (replacement) toxicity testing methods that are now available have been developed largely due to the phased EU ban on the animal testing and marketing of cosmetic ingredients and finished products. As a result, discussions on the replacement of animals in toxicity testing are far more common and perhaps are considered more acceptable in the cosmetics field. However, young researchers working or studying in other areas of toxicity may feel less able to speak out about their research interests, especially if they involve replacement/non-animal methods, as these are seen as more controversial than ‘two Rs’ (reduction or refinement) approaches. It is therefore important to recognise that these methods are now of essential use in other chemical testing sectors, such as the food or pharmaceutical industries. This demonstrates that when a non-animal method is developed and accepted, it can potentially be applied to the testing of any substance, for any purpose. This may encourage young researchers to voice their interests in the development and use of non-animal methods.
This is especially relevant as, despite the development of alternatives for use in areas such as cosmetics, toxicity testing in animals continues in many other industries. For example, in the UK in 2013, over 375,000 toxicity tests (from a total 4.12 million procedures) were performed on animals (mice, rats, rabbits, guinea-pigs, dogs, cats, monkeys, birds and fish).12 Another important point with regard to the Young Researcher Lush Prize, is that almost half of all animal experiments in the UK are carried out at  universities. One of the most concerning findings is that the increase in the use of genetically-modified animals (mainly mice) and the increasing use of zebrafish are, in some contexts, being considered as ‘alternatives’. This was highlighted by FRAME in a report on the Home Office annual statistics on animal use in Great Britain in 2012.13
The 2011 EU figures14 showed that over 1 million animals (1,004,873, from a total of just under 11.5 million animals) were used in toxicity testing across the EU states in that particular year. Of these, 111,166 animals were used in tests that were not even required by law (categorised as ‘no regulatory requirements’). The  archaic and much criticised LD50/LC50 (lethal dose or lethal concentration test, which tests the amount of substance required to kill 50% of the animals tested) accounts for the majority of the animals used each year, along with other lethal tests (34%). The other main use is simply categorised as ‘other’ toxicology tests (22%), followed by chronic/sub-chronic toxicity and reproductive toxicity. There is no official figure for the number of toxicity tests still conducted on animals worldwide (from the estimated yearly total of 115 million animals15 used in all experiments), as many countries omit this information or do not even count the numbers of animals used. However, a revised  estimate by Lush Prize researchers puts the number of toxicity tests carried out on animals worldwide at almost 9.5 million (from a total 118 million animal experiments).16

After the 2013 marketing ban, much work is still to be done

Although the EU cosmetics legislation has been the major driver of the development of non-animal toxicity testing methods in recent years, there is still much more to be done. This is illustrated very clearly, given that “Over 80% of the world allows animals to be used in cruel and unnecessary cosmetics tests and these animal tested cosmetics can be purchased in every country across the globe.”17
The proposed 2013 EU marketing ban on animaltested  cosmetics did finally go ahead, though the European Commission had previously considered the possibility of delaying the deadline on the basis of recommendations that necessary but still missing’ alternative methods would take much longer to be developed. For example, estimates of another 5–9 years were proposed for methods for skin sensitisation and toxicokinetics to be developed, and possibly even longer for full replacement in these areas. No estimates were provided at all for when repeat-dose toxicity, reproductive toxicity or carcinogenicity tests on animals might be developed. These timelines were estimated in a report published by the Commission in 2011.18  In the three years since that time, aside from the introduction of the 2013 ban itself (which went ahead regardless of the lack of alternatives available, which was great news), further work had been ongoing in the areas of toxicity testing which still need development. For example, in 2013, the Joint Research Centre (JRC) published its EURL-ECVAM Strategy to Avoid and Reduce Animal Use in Genotoxicity Testing.19 Similarly, the five-year long NOTOX project,20 launched in 2011 and involving a network of scientific expertise from several countries, works “towards the replacement of current repeated dose systemic toxicity testing in human safety assessment”. NOTOX is part of a wider project , funded under the EU Seventh Framework Programme (FP7), known as SEURAT (Safety Evaluation Ultimately Replacing Animal Testing). This project combines the research efforts of over 70 European universities, public research institutes and companies, and regularly posts open vacancies and research opportunities.21 Of particular relevance is that SEURAT hosted a Young Scientists Summer School to discuss replacement of repeat-dose toxicity testing in animals.22

Focusing on key areas

As highlighted by EURL-ECVAM,23 a key factor in the development of non-animal methods is the integration of a number of alternative test methods into a ‘battery’ that successfully addresses a number of endpoints, especially those which are considered more complex or need to be considered in-depth. For example, several alternatives available for skin testing, examine how a substance may react in various stages of topical application, absorption, irritation or corrosion, and provide very targeted and quantitative results, especially when compared to a crude skin test in rabbits or guinea-pigs. Therefore, non-animal methods which are still under development or undergoing validation, or ‘gaps’ in the development of methods where the greatest use of animals still occurs, such as reproductive or chronic toxicity, could be helped by awarding the Lush Prize to young researchers to allow them to potentially channel their ideas or research themes for specific replacement projects and encourage them to specialise in key areas.
Never before has there been a better opportunity for young researchers to focus on replacement alternatives, not only to save animals from unnecessary pain and suffering, but also to pave the way for a career in cutting-edge innovation. As highlighted by the New England Anti-Vivisection Society (NEAVS) in a previous interview with Lush Prize, linking an early-career scientist’s research to increased income and sponsorship is key:
“I think the solution for graduate students who want  to do more progressive in vitro research is to find the granting agencies that will help bring money in [to an institution]. …The key to changing institutions is bringing in grant dollars. When someone who wants to develop in vitro alternatives can show that they can bring in million-dollar grants, then institutions are going to have to accept it. They’re not going to turn money away, even if they want to try to suppress a certain ideology”.24
What this means, in effect, is that, if a researcher has ideas, but can also say “if you fund me, I propose to cut your costs, save you time, increase income and improve your business”, whilst this might be viewed as a challenge, their proposals are much more likely  to be considered.

Challenges for ethical early-career scientists

As previously highlighted, there remain ongoing prejudices toward switching from animal to non-animal research. Resistance to change, combined with ‘comfort’ in repeating accepted, conventional  methods, allows the animal research industry to maintain the status quo, despite ever-increasing recognition that animal testing is a flawed, overrated and outdated system. It must also be noted that the industry has, for decades, consisted of a network, not only of researchers, but also breeders, suppliers and transporters of animals across the world, who rely on animal testing to continue. There are other factors to consider — for example, some scientists (especially senior-level researchers) have based their entire careers on the use of animals, and are unable or unwilling to consider anything else; they may view switching to non-animal research as  the daunting and unattractive option of ‘starting again’. This may also apply to earlier career individuals, who have followed the mainstream route into animal-based toxicology to progress their careers to date, for example, since leaving university. This is echoed by several previous prizewinners, who felt that the undergraduate level of their education was the most challenging arena in trying to avoid the use of animals. Nevertheless, one positive finding from interviews with previous Young Researcher Prize winners is that “Young scientists don’t always have the prejudices about animal testing being the ‘best’ way of doing things”.2
Finally, to provide some useful insight into the types of scholarships available to young researchers, Appendix 1 gives a summary of 15 PhD studentships recently funded by the UK NC3Rs. A full list is shown to illustrate the types of research being undertaken
— however, it must be noted that the studentships cover the broader remit of the Three Rs, rather than the ‘replacement only’ criterion that the Young Researcher Prize demands.

Download the full article here (including Appendix 1) 

Rebecca Ram
Lush Prize
ECRA
41 Old Birley Street
Manchester M15 5RF
UK
E-mail: rebecca@lushprize.org

References and Notes

1 Russell, W.M.S. & Burch, R.L. (1959). The Principles of Humane Experimental Technique, 238pp. London, UK: Methuen.
2 Anon. (2012). Lush Young Researchers Prize 2012 — Research Paper, 22pp. Available at: http://www. lushprize.org/wp-content/uploads/2012/06/Lush-
Young-Researchers-Prize-2012-Research-Paper.pdf (Accessed 05.11.15).
3 Anon. (2014). In-Vitro Toxicology Testing Market worth $17,227 Million by 2018. Available at: http://www. prnewswire.co.uk/news-releases/in-vitro-toxicologytesting- market-worth-17227-million-by-2018-25358636 1.html (Accessed 05.11.15).
4 Interview with PETA, 20 August 2012.
5 Willetts, D. (2014). Hansard Written Answers. Animal Experiments: Business, Innovation and Skills written question — answered on 11th March 2014. Available at: http://www.theyworkforyou.com/wrans/?id= 2014-03-11a.188641.h (Accessed 05.11.15).
6 NC3Rs (2014). Funding schemes. Available at: http:// www.nc3rs.org.uk/landing.asp?id=27 (Accessed 06. 06.14).
7 NC3Rs (2014). Quantifying the behavioural and facial correlates of pain in laboratory macaques. Available at:  https://www.nc3rs.org.uk/quantifyingbehavioural-
and-facial-correlates-pain-laboratorymacaques
(Accessed 05.11.15).
8 NC3Rs (2012). The Rabbit Grimace Scale — A new method for pain assessment in rabbits. Available at: https://www.nc3rs.org.uk/news/rabbit-grimacescale-% E2%80%93-new-method-pain-assessment-rabbits
(Accessed 05.11.15).
9 Taylor, K. (2014). EU member state government contribution to alternative methods. ALTEX 31, 215– 218.
10 Anon. (2013). Europe Bans Marketing of Cosmetics Tested on Animals. Available at: http://ensnewswire. com/2013/03/11/europe-bans-marketing-o
f-cosmetics-tested-on-animals/ (Accessed 05.11.15). 11 Schwarz, M. (2011). Meta Analysis of a Battery Test of Reproductive Toxicity Assays: The ReProTect Experience. [Presentation given at Open Tox, Munich,
9–12 August, 2011.] Available at: http://www.opentox.org/meet/opentox2011/talks/OpenTox2011_
Talk-Schwarz.pdf (Accessed 05.11.15).
12 Home Office (2013). Annual Statistics of Scientific Procedures on Living Animals — Great Britain 2013, 59pp. London, UK: Her Majesty’s Stationery Office. Available at: https://www.gov.uk/government/ uploads/system/uploads/attachment_data/file/3278 54/spanimals13.pdf (Accessed 05.11.15).
13 Hudson-Shore, M. (2013). Statistics of Scientific Procedures on Living Animals 2012: Another increase in experimentation — Genetically-altered animals dominate again. ATLA 41, 313–319.
14 Anon. (2013). Report from the Commission to the Council and the European Parliament. Seventh Report on the Statistics on the Number of Animals Used for Experimental and Other Scientific Purposes in the Member States of the European Union. COM (2013) 859 final, 14pp. Brussels, Belgium: European Commission. Available at: http://eur-lex.europa.eu/
legal-content/EN/TXT/PDF/?uri=CELEX:52013DC085
9&from=EN (Accessed 05.11.15).
15 Taylor, K., Gordon, N., Langley, G. & Higgins, W. (2008). Estimates of worldwide laboratory animal use in 2005. ATLA 36, 327–342.
16 Anon. (2014). The 2014 Lush Prize: A Global View of
Animal Experiments 2014, 42pp. Available at: http://
www.lushprize.org/wp-content/uploads/Global_
View_of-Animal_Experiments_2014.pdf (Accessed 05.11.15).
17 Cruelty Free International (2012). Did you know animal tested cosmetics are for sale in every country in the world? Available at:  ttp://www.crueltyfree
international.org/en/the-issue (Accessed 06.06.14).
18 Adler, S., Basketter, D., Creton, S., Pelkonen, O., van Benthem, J., Zuang, V., Andersen, K.E., Angers- Loustau, A., Aptula, A., Bal-Price, A., Benfenati, E.,
Bernauer, U., Bessems, J., Bois, F.Y., Boobis, A., Bran – don, E., Bremer, S., Broschard, T., Casati, S., Coecke, S., Corvi, R., Cronin, M., Daston, G., Dekant, W., Felter, S., Grignard, E., Gundert-Remy, U., Heinonen, T., Kimber, I., Kleinjans, J., Komulainen, H., Kreiling, R., Kreysa, J., Leite, S.B., Loizou, G., Maxwell, G., Mazzatorta, P., Munn, S., Pfuhler, S., Phrakonkham, P., Piersma, A., Poth, A., Prieto, P., Repetto, G., Rogiers, V., Schoeters, G., Schwarz, M., Serafimova, R., Tähti, H., Testai, E., van Delft, J., van Loveren, H., Vinken,
M., Worth, A. & Zaldivar, J.M. (2011). Alternative (nonanimal)
methods for cosmetics testing: Current status and future prospects — 2010. Archives of Toxicology 85, 367–485.
19 Corvi, R., Madia, F., Worth, A. & Whelan, M. (2013). EURL ECVAM Strategy to Avoid and Reduce Animal Use in Genotoxicity Testing, 48pp. Ispra, Italy: European Commission, Joint Research Centre, Institute for Health and Consumer Protection. Available at: http://publications.jrc.ec.europa.eu/repository/bitstream/
111111111/30088/1/jrc_report_en_34844_on line.pdf (Accessed 05.11.15).
20 NOTOX (undated). Welcome to NOTOX. Available at: http://www.notox-sb.eu/ (Accessed 05.11.15).
21 SEURAT (undated). Welcome to the SEURAT-1 website. Available at: http://www.seurat-1.eu/ (Accessed 05.11.15).
22 SEURAT (undated). SEURAT-1 & ESTIV Joint Summer School — 8–10 June 2014. Egmond aan Zee, Netherlands. Available at: https://www.eurtd.com/
seurat-1/2014/summer-school/ (Accessed 05.11.15).
23 Anon. (2013). EURL ECVAM Progress Report on the Development, Validation and Regulatory Acceptance of Alternative Methods (2010–2013). Available at: http://ihcp.jrc.ec.europa.eu/our_labs/eurl-ecvam/
eurl-ecvam-releases-2013-progress-reportdevelopment-
validation-regulatory-acceptancealternative- methods (Accessed 06.06.14).
24 Interview with NEAVS, 20 August 2012.

On Replacing the Concept of Replacement

Michael Balls

Russell and Burch saw failure to accept the correlation
between humanity and efficacy as an example of rationalisation,
a psychological defence mechanism

While wondering what I could discuss in this column I looked, as I often do, in the abridged version1 of The Principles of Humane Experimental Technique,2 at Russell and Burch’s introduction of what I call the humanity criterion. It is part of their discussion of the sociological factors which are among the Factors Governing Progress. This is how part of page 101 of the abridged version reads:

In fact, really informative experiments must be as humane as would be conceivable possible, for science and exploration are indissolubly linked to the social activity of cooperation, which will find its expression in relation to other animals, no less than to our fellow humans. Conscious good will and the social operational method are useless as safeguards against the mechanism of rationalisation (in the pathological sense of the term – i.e. the mechanism of defence by which unacceptable thoughts or actions are given acceptable reasons to justify them to oneself and to others, while, at the same tie, unwittingly hiding the true, but unconscious, motives for them).

The bold type indicates my explanation, and I have to admit that, six years after preparing the abridged version of The Principles, I now found it difficult to see what Russell and Burch had intended to convey. I therefore looked back at the original book, and found this paragraph on pages 156−157:

In efficacy, or yield of information, the advantages of humane technique apply almost universally. The correlation between humanity and efficacy has appeared so often in this book that we need not labour the point. There is, however, a more fundamental aspect of this correlation, specially important in research. Science means the operational method — telling somebody else how to see what you saw. This method is one of the greatest of all human evolutionary innovations. It has, however, one drawback. It prevents permanent acceptance of false information, but it does not prevent wastage of time and effort. The activity of science is the supreme expression of the human exploratory drive, and as such it is the subject to the same pathology. The scientist is liable, like all other individuals, to block his exploration on some front where his reactions to childhood social experiences are impinged upon. When this happens to the experimental biologist, we can predict the consequence with certainty. Instead of really exploring, he will, in his experiments, act out on his animals, in a more or less symbolic and exaggerated way, some kind of treatment which he once experienced in social intercourse with his parents. He can rationalise this as exploration, and hence fail to notice the block. But in fact such acting out invariably occurs precisely when real exploration is blocked, and must be relinquished before real exploration can begin again. Hence, such experiments will be utterly wasteful, misleading, and uninformative. The treatment of the animals, for one thing, will inevitably be such as to impair their use as satisfactory models. The interpretation of the results will be vitiated by projection. Really informative experiments, must in fact be as humane as would be conceivably possible, for science and exploration are indissolubly linked to the social activity of cooperation, which will find its expression in relation to other animals no less than to our fellow humans. Conscious goodwill and the social operational method are useless as safeguards against the mechanism of rationalisation (in the pathological sense of the term).

Here, the underlining indicates what I omitted from the abridged version, and I now wonder why I did so. These words clearly reflect Russell’s interest in psychology — he later became a psychotherapist, and undoubtedly will have been influenced by discussions with his psychotherapist wife, Claire Russell. They could be seen as an explanation why some scientists did not appreciate the essential link between humanity and efficacy, and why Russell thought they needed what was offered by the Three Rs and the humanity criterion.

It is not clear what is meant by “the social operational method”, and consulting Google leads to only one hit — The Principles itself! “Conscious goodwill” is probably meant to contrast with unconscious rationalisation.  Perhaps what Russell meant is that, however sincere the intention may appear to be, support for the Three Rs is useless, unless it leads to active and practical commitment to their development and application.

We are often confronted with rationalisation, the pseudo-rational justification of irrational acts,3 and its relative, intellectualisation, a different defence mechanism (or way of making excuses), “where reasoning is used to block confrontation with an unconscious conflict and its associated emotional stress, where thinking is used to avoid feeling. It involves removing one’s self, emotionally, from a stressful event. Intellectualisation is one of Freud’s original defence mechanisms. Freud believed that memories have both conscious and unconscious aspects, and that intellectualisation allows for the conscious analysis of an event in a way that does not provoke anxiety.”4

I am not a psychoanalyst, and I think it would be unwise, even dangerous, were I to seek to delve into the underlying reasons why some scientists are so keen to run to animal experimentation as the first resort and to do so little to make possible its replacement. Nevertheless, I can say, without fear of contradiction, that this is another great example of how Russell and Burch’s wonderful book continues to give us food for thought and calls for action.

Professor Michael Balls
E-mail: michael.balls@btopenworld.com

References
1 Balls, M. (2009). The Three Rs and the Humanity Criterion, 131pp. Nottingham, UK:  FRAME.
2 Russell, W.M.S. & Burch, R.L. (1959). The Principles of Humane Experimental Technique, xiv + 238pp. London, UK: Methuen.
3 Anon. (2015). Rationalization (psychology). San Francisco, CA, USA: Wikipedia Foundation, Inc. Available at: https://en.wikipedia.org/wiki/Rationalization_(psychology) (Accessed 26.08.15).
4 Anon. (2015). Intellectualization. San Francisco, CA, USA: Wikipedia Foundation, Inc. Available at: https://en.wikipedia.org/wiki/Intellectualization (Accessed 26.08.15).

The Principles of Humane Experimental Technique is now out of print, but the full text can be found at http://altweb.jhsph.edu/pubs/books/humane_exp/het-toc. The abridged version, The Three Rs and the Humanity Criterion, can be obtained from
FRAME.

Download a pdf copy of this post by clicking here.

Previous Wisdom of Russell & Burch posts from Michael Balls:

The Concept, Sources and Incidence of Inhumanity and its Diminution or Removal Through Implementation of the Three Rs.
The Wages of Inhumanity.
Fidelity and Discrimination.
Reduction.
Refinement. 
Replacement. 
The Factors Governing Progress. 
UFAW and Major Charles Hume. 
The Toxicity Testing Problem. 
The Use of Lower Organisms.
The Analysis of Direct Inhumanity.
William Russell: Polymath, Wordsmith, Classicist and Humourist .
Rex Leonard Burch: Humane Scientist and Gentle Man.
On the Proper Application of Appropriate Statistical Methods. 
Comparative Substitution. 
The Three Rs: The Way Forward .
The Choice of Procedures .
Rationalisation and Intellectualisation.

 

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
benefits.

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
BASF SE
Experimental Toxicology and Ecology
Ludwigshafen am Rhein
Germany
E-mail: robert.landsiedel@basf.com

Dr Wera Teubner
BASF Schweiz AG
Product Safety
Basel
Switzerland

References

1 ECHA (2015). Read-Across Assessment Framework (RAAF) [ECHA-15-R-07-EN], 38pp. Helsinki, Finland: European Chemicals Agency. Available at: http://
echa.europa.eu/documents/10162/13628/raaf_en.pdf
(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.
edu/pubs/books/humane_exp/het-toc
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.
europa.eu/documents/10162/13634/echa_jrc_sla_
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,
586–601.
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–
253.
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://
echa.europa.eu/documents/10162/13632/information_
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: http://echa.europa.eu/documents/ 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: http://echa.europa.eu/documents/10162/
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: http://www.oecd.org/chemicalsafety/riskassessment/
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: http://www.echemportal.org/echemportal/
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. europa.eu/information-on-chemicals/registeredsubstances (Accessed 12.11.15).
30 SEURAT (undated). Welcome to the SEURAT-1 website. Available at: http://www.seurat-1.eu/ (Accessed 11.11.15).
31 EPA (2015). Toxicity Forecasting: Advancing the Next Generation of Chemical Evaluation. Durham, NC, USA: US EPA Research. Available at: http://www2.epa.gov/
chemical-research/toxicity-forecasting (Accessed 12.11.
15).

Download the pdf here

A new approach to optimise the use of animal models in drug discovery through Big Data Sharing

Jiaqi Lu and Jianfei Wang

Sharing full details of animal models would enhance
consistency of models between establishments and
reduce the numbers of animals used for set-up and validation

With the continued rapid growth of Information technology (IT) and Cloud technology, the  concept of so called ‘Big Data’ has emerged over the last decade. Right now, this buzzword-phrase is being heard and talked about much more in the pharmaceutical industry and health care sectors.1–3 As a result of increased investment and better supply of information, pharmaceutical companies have recently been collating years of research and development data, including in vitro and animal preclinical test information, into medical databases. Meanwhile, the governments and other public stakeholders have been opening their vast stores of health-care knowledge, including data from clinical trials and information on patients acquired through public medical insurance programmes. 4,5 In parallel, some cross-boundary large IT companies, such as Google and Apple, have also jumped into this hot-tub and expect financial returns from the accelerating value and innovation in the health care and drug discovery industries.6,7

Historically, drug discovery and development has been a relatively isolated endeavour, with little information sharing evident among different pharmaceutical companies and academic researchers. In recent years, however, it has become apparent that pharmaceutical R &D is suffering from declining success rates and stagnant pipelines. This provides the impetus and an opportunity to change the landscape of drug discovery by utilising Big Data information. The current ability to generate and store vast amounts of information has led to an abundance of data and a growth in the discipline of Systems Pharmacology. Data are generated from several stages in the drug discovery process, including pre-clinical animal tests and Phase I, II and III trials, as well as post-marketing monitoring.8 Effectively utilising these data will help pharmaceutical companies to better identify new potential drug candidates and to develop safe, effective, approved medicines more quickly. In this article, a new approach to optimising the use of animal models in drug discovery through Big Data is explored.

Challenges and opportunities

Economic pressures, perhaps more than any other factor, are driving the demand for Big Data analysis and applications in drug discovery; this is appropriate, as it costs more than one billion US dollars to test and develop one new drug, and it often takes far more than ten years. In the early pre-clinical stage, the increasing costs of high-throughput screening of compound candidates, and of safety and efficacy studies in animal models, are major financial challenges to all pharmaceutical companies.

Animal model data are an important part of drug discovery Big Data — however, primary data generated from drug discovery animal research are infrequently accessed or re-used remotely from where they were generated. There is limited access to detailed drug discovery animal model data, since full information about the protocols has not always been published in research papers. This has led to a drive by many journals to enhance the depth of details given in the Methods sections  of manuscripts. Without this attention to detail, there have been missed opportunities for the continuous optimisation and  improvement of scientific methods and enhancement of innovation.

On the other hand, with the strengthening social pressures to avoid the use of laboratory animals in drug discovery, pharmaceutical companies and academia are finding it hard to demonstrate the application of the Three Rs principles to the satisfaction of the public. The sharing of full details of the animal models used in drug discovery would enhance the consistency of models between establishments and reduce the numbers of animals used to set up and validate the models.

Last, but not least, changing the mind set about confidentiality is a big challenge for all public and private organisations. Pharmaceutical R&D has always been a ‘secretive’ activity, conducted within the confines of the R&D department, with little external collaboration. Unless it is possible to identify an ideal future state with non-competitive aims, there is little value to investing in improving Big Data sharing capabilities. Today, public–private partnerships still represent a concept to be tested — therefore, if a new approach of sharing the full details of animal models used in drug discovery demonstrated its value and reduced attrition, then there would be an enlarged space for future developments in sharing information.

Landscape and strategy

The sharing and cross-analysis of pre-competitive drug discovery animal model information across public research organisations, pharmaceutical companies, Three Rs organisations, biotech companies and contract research organisations (CROs), through a one-stop sharepoint, would contribute to the simplification of the partners’ operating systems, would facilitate the delivery of more products of value through reduced attrition, and would enable all the partners to build trust by demonstrating their commitment to the Three Rs.

This one-stop sharepoint would allow data circulation within and beyond the original partnership. By enhancing interdisciplinary scientific reviews, animal studies could be optimised. Raw data, including cross-therapeutic animal models and protocols, drugvehicle effects, and positive and negative study results, would permit the assessment of the positive predictive value of each animal model. In addition, new information and hypotheses generated from data cross-analysis could be available to all partners, which would maximise the value of the animal research data. This non-competitive animal model information, such as guidelines on contemporary best practice and innovative alternative approaches to animal research, could also contribute to public knowledge and enhance animal welfare. In the end, by exchanging this information, all the partners would bolster external collaborations within and beyond the original partnership (Figure 1).
Figure 1

The importance of partnerships

The key component to achieving this goal is through partnership. No matter how public  research organisations, pharmaceutical companies, Three Rs organisations, CROs and biotech companies break the silos by enhancing collaboration with external partners,
all stakeholders can extend their knowledge and data networks through partnership.

Ideally, the following objectives could be achieved: identifying and discontinuing the use of animal models that are not sufficiently robust or fail to translate in the clinic; optimisation of the design and validation of animal models and protocols (e.g. to improve translation of animal models between laboratories or decrease model severity, to facilitate the informed choice of animal models, to share best practices and Three Rs advances, and to reduce duplication of efforts).

Academic partners could share insights from the latest scientific breakthroughs in cross-therapeutic animal models and make a wealth of innovation available. Normally, academia is willing to help improve the transfer of animal models between laboratories and the intra-laboratory reproducibility. Collaborations between the pharmaceutical companies could quickly identify and discontinue the use of animal models that identify treatments that fail to translate into efficacious medicines in the clinic. This could then lead to optimising the design/validation of animal models and protocols as a next step. This partnership could reduce clinical attrition, which would, in turn, reduce the financial cost of whole drug discovery process. Through collaboration with Three Rs organisations in order to learn best practices and Three Rs advances, stakeholders would enhance animal welfare by direct innovation and implementation of non-animal alternatives where they are the most needed, or by refining animal models to decrease severity and variability. Maximising external collaborations with CROs and new biotech companies could quickly add to or scale up internal capabilities and provide access to expertise in advanced technologies and animal models which would otherwise require establishment in-house.

Although this pattern of collaboration appears to be a win–win situation for all the partners, belief in the benefits of the data sharing culture and active participation still needs to be inspired. All stakeholders would have to recognise the value of Big Data analysis and sharing and be willing to act on its insights; a fundamental mind-set shift for many and one that may prove difficult to achieve. Confidentiality issues would also continue to be a major concern, although new IT technology can readily enhance private information protection in the databases. However, stakeholders would still have to be vigilant and watch for potential problems, as the increasing amount of information on animal models that is becoming openly available has the potential to the misunderstood by the general public.

Perspective

Big Data sharing is a new approach to optimising the use of animal models in drug discovery. Sharing animal model information, such as protocols, study results (including drug-vehicle effects and positive and negative data) and translational outcomes, in a single cross-therapeutic platform that uses a standard data capture and common ontology framework, would permit the secondary analysis of multiple datasets.

This would lead to higher efficiency for the assessment of preclinical drug efficacy and pharmacokinetics, and would also reduce the welfare impact on animals. Ultimately, it will also facilitate the re-use of animal data in the wider and more complex scenario of drug R&D, by facilitating linkage with other datasets, such as safety assessment datasets, chemistry and pharmacokinetics and pharmacodynamics (PK/PD). Crosspharma
collaborations in animal research are identified as a high-impact opportunity for accelerating scientific innovation and improving scientific output in animal model research for drug discovery, and for more tangible contributions to the Three Rs ethical principles across all the pharmaceutical industry. Subsequently, animal test dataset sharing across multiple pharmas and some prominent CROs, would further permit the appropriate assessment of the value of animal use in drug discovery and would lead to a reduction in the numbers of animals used in this work. Finally, if the pool of animal datasets generated by pharmas/CROs were ultimately augmented by the experimental data from many prominent academic institutions, it would be possible to generate an animal test search engine similar to Google Scholar. In an ideal situation, any proposal to carry out an animal test or use a particular animal model, should start with a search to identify suitable assays and an assessment of the potential utility of the model. This could be done by viewing existing positive and negative animal test data, as well as easily contacting partners experienced in these assays for advice. In addition, further animal model optimisation could be performed or unnecessary animal tests could be prevented. This would ultimately reduce animal use and reduce drug discovery costs and would speed up the drug discovery process by affording a greater chance of successful translation of efficacy to the clinic.

Acknowledgments

This work was supported by Key Projects in the National Science & Technology Pillar Program (No.  2011BAI15B03). The authors wish to thank Dr David Tattersall and Cheng Gao for their valuable suggestions.
Dr Jiaqi Lu
Department of Laboratory Animal Sciences
GlaxoSmithKline, R&D China

Author for correspondence:
Dr JianFei Wang
Head, Laboratory Animal Sciences
GlaxoSmithKline, R&D China
2F, Building 3
898 Halei Road
Zhangjiang Hi-Tech Park
Pudong
Shanghai 201203
China
E-mail: jianfei.j.wang@gsk.com

References

1 Fabricio, F.C. (2014). Big data in biomedicine. Drug Discovery Today 19, 433–440.
2 Marx, V. (2013). The big challenges of big data. Nature, London 498, 225–260.
3 Groves, P., Kayyali, B., Knott, D. & Kuiken, V.S. (2013).   The ‘big data’ revolution in healthcare: Accelerating value and innovation, 19pp. Center for US Health System Reform, Business Technology Office, McKinsey & Company.
4 SOTP (2012). Obama Administration Unveils “Big Data” Initiative: Announces $200 Million in New R&D Investments, 4pp. Washington, DC, USA: Office of Science & Technology Policy, Executive Office of the President, White House. Available at: https://www. whitehouse.gov/sites/default/files/microsites/ostp/big _ data_press_release_final_2.pdf (Accessed 19.09.15).
5 NIH (2015). NIH-led effort launches Big Data portal for Alzheimer’s drug discovery. Bethesda, Maryland, USA: National Institutes of Health. Available at: http://
www.nih.gov/news/health/mar2015/nia-04.htm (Accessed 19.09.15).
6 Harris, D. (2015). Google, Stanford say big data is key to deep learning for drug discovery. Houston, TX, USA: Knowingly, Corp. Available at: https://gigaom.com/ 2015/03/02/google-stanford-say-big-data-is-key-todeep-learning-for-drug-discovery/ (Accessed 19.09.
15).
7 Anon. (2015). Apple HealthKit. Cupertino, CA, USA: Apple Inc. Available at: https://developer.apple.com/ healthkit/ (Accessed 19.09.15).
8 Zhang, J., Hsieh, J.H. & Zhu,  H. (2014). Profiling animal toxicants by automatically mining public bioassay data: A Big Data approach for computational toxicology.
PLoS One 9, e99863.

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

Overview

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.

Limitations

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
Setagaya-ku
Tokyo 158-8501
Japan
E-mail: h-kojima@nihs.go.jp

References

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: http://www.mattek.com/ (Accessed 30.07.15).
4 Xenometrix (2015). Homepage. Allschwil, Switzerland: Xenometrix AG. Available at: http://www.xenometrix.ch/ (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: http://anzeninfo.mhlw.go.jp/user/anzen/kag/ankg_ghs.htm (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: http://www.oecd-ilibrary.org/environment/oecdguidelines-for-the-testing-of-chemicals-section-4-
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: http://www.oecd-library.org/environment/testno-430-in-vitro-skin-corrosion-transcutaneouselectrical-resistance-test-method-ter_9789264242 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: http://www.oecd-ilibrary.org/environment/testno-431-in-vitro-skin-corrosion-reconstructed-humanepidermis-rhe-test-method_9789264242753-en; 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: http://www.oecd-ilibrary.org/environment/testno-
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:  http://www.fda.gov/downloads/drugs/guidance 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: http://www.oecd-ilibrary.org/environment/testno-492-reconstructed-human-cornea-like-epitheliumrhce-test-method-for-identifying-chemicals-not-requiring-classification-and-labelling-for-eye-irritation-orserious-eye-damage_9789264242548-en;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: http://www.fda.gov/ 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.

Development and Validation of a Low-fidelity Simulator to Suture a Laparotomy in Rabbits

Juan J. Pérez-Rivero, Tonantzin Batalla-Vera and Emilio Rendón-Franco

An easily constructed, low-cost simulator
is assessed for its efficacy in the surgical training
of veterinary science undergraduates

Download a pdf of this article

Introduction

There is a growing need for the development of alternatives to reduce, replace and refine  the use of animals for surgical training in contemporary veterinary education at the undergraduate level. In the present study, a simulator to suture a midline laparotomy in the rabbit was designed, that could be constructed from widely-available and low-cost materials. The simulator was used to develop surgical skills in students at the undergraduate level of veterinary medicine. Thirty-five, third-year veterinary students, with no previous surgical experience, were divided into two groups: a control group that did not use the simulator (n = 19), and an experimental group that used the simulator three times to practise the suturing of a laparotomy (n = 16). Later, both groups performed
a midline laparotomy in an anaesthetised rabbit, and the rate of closure of each anatomical plane (peritoneum, additional reinforcement, and skin) was measured.

The usefulness of simulators

The surgical training of undergraduate students by using live animals provides few opportunities for real training and is applicable only to certain surgical techniques. In addition, it also raises serious ethical and animal welfare considerations. The students themselves are also subjected to a level of stress, this being, in most cases, a cause of errors. Consequently, they do not adequately benefit from the training provided.1,2

In veterinary medicine and animal sciences, the Three Rs principles are being implemented as widely as possible. This involves the reduction, replacement and refinement of animal use, both in experiments and in teaching.3 One way to accomplish this is through the use of various simulators in their different forms, such as synthetic simulators, multimedia simulations, virtual reality, carcasses, and ethically sourced animal tissues.4,5 These provide training alternatives, which permit the acquisition of skills to successfully meet the needs of future clinical and surgical experiences with live patients, and to ensure that maximum educational value is achieved during practical training.6

The fidelity of a simulator is determined by how much realism is provided through characteristics such as visual cues, touch, the ability to feedback, and interaction with the student. In general, simulators can be divided into two groups: high-fidelity simulators, which are usually highly technical, detailed and realistic; and low-fidelity simulators, which have a low level of realism, are usually made with widely available and low-cost materials, are often portable, and can be used on a table. Despite their simplicity, the latter group of simulators assist the development of psychomotor skills.7 Some limitations of the use of simulators are related to their cost or difficulty in sourcing spare parts. Moreover, despite the large number of simulators that have been developed, few studies have been conducted to evaluate their effectiveness, leaving the concept of teaching through simulators at an empirical stage.8 Therefore, it is necessary to develop inexpensive, easy-to-construct simulators that support the process of surgical teaching, and also to quantitatively assess the effectiveness of their use. Therefore, the aim of this work was to develop and validate a low-fidelity simulator to assist in the teaching of the correct technique for closing a rabbit midline laparotomy.
Simulator assembly

Development of the simulator

A 10cm long and 4cm wide opening was made in an empty plastic 500ml solution bottle (Pisa Agropecuaria, Guadalajara, Jalisco, Mexico), leaving protruding areas to represent both the xiphoid process and the pubic symphysis (Figure 1a). To give support to the bottle, an internal  cardboard lining was added, as well as three 3ml syringes widthwise (Figure 1b). Two 3mm thick silicone sheets were made by pouring 270ml of PE53® silicone rubber (Poliformas Plasticas, Mexico City, Mexico) into a mould, 23cm long by 13cm wide, which was allowed to set at room temperature for 24 hours.

The back-board from a standard paper clipboard was used as the simulator base, with the plastic bottle placed onto the board and the first sheet of silicone overlaid, in order to simulate the peritoneum (Figure 1c). Subsequently, the rectus abdominis muscles were simulated by placing two sheets of 3mm thick × 28cm long × 21cm wide polypropylene around the bottle (Foamy; Mylin, Mexico City, Mexico), leaving a gap of 3cm in width along the entire midline. Finally, this layer was covered with the second sheet of silicone to simulate skin, and both sheets of silicone were tightened onto the clipboard base with paper clips (Figure 2a). The appropriate size head, thorax and abdominal organs were fashioned from cotton fabric and added to the simulator prior to use (Figure 2a).
General view

Simulator validation

Thirty-five students in the third year of a veterinary medicine and zootechnics course at the Universidad Autónoma Metropolitana, Unidad Xochimilco, with no previous experience in surgery, received a 120-minute theory session, supported with slides, on the midline laparotomy technique and suture in rabbits.9 This was part of the Surgical-Veterinary Therapeutic Bases module. Later, the students were divided into two groups: the experimental group (n = 16), which  was organised in four surgical teams of four participants each, and the control group (n = 19), which was divided in four groups of four participants and one three-participant group. Each student was assigned his/her rotation within the group, in such a way that they all covered all the roles once (surgeon, first assistant, scrub nurse, and anaesthesiologist). Each surgeon/first assistant team (according to the assigned rotation) of the experimental group used the laparotomy simulator two days prior to the practice on the live animals. They were asked to repeat three times the following procedure: put the surgical drapes in place (Figure 2b); perform a 7cm incision, including all the layers of the simulator; suture, with continuous stitches, the first silicone layer (peritoneum), which was reinforced with inverted ‘U’ stitches; and suture, with Sarnoff stitches, the second silicone layer (skin). The first assistant was only allowed to help the surgeon in handling the surgical instruments that were used. The closure of planes was performed by using nylon 2-0 suture (Figure 3).
Simulator in use

Subsequently, the participants of both groups performed midline laparotomies on 35 clinically healthy New Zealand rabbits (Oryctolagus cuniculus), suturing midline (peritoneum) with continuous stitching, reinforcing (muscular fascia) with inverted ‘U’ stitches, and suturing the skin with Sarnoff stitches, all performed under general anesthesia, according to the method previously described by Perez-Rivero and Rendón-Franco.10 Both the control group and the experimental group performed one surgery weekly. In total, evaluations were completed in 4 weeks (i.e. one week for each participant from each team).

Since the lengths of the incisions were different in all the cases, the rate of closure of each anatomic plane and all planes in total, was calculated as follows: the length of each incision was measured (in centimetres), and this was divided by the time (in minutes) taken to complete the suturing. The result was expressed in minutes per linear centimetre of incision (MLCI). During the whole process, each group was supervised by two professors and five assistant instructors.
Table 1

Statistical Analysis

Students having the prior role of first assistant, scrub nurse, and/or anaesthesiologist, would have previously observed and/or helped in the performance of the laparotomy. This could have resulted in an improvement in their performance when participating as actual surgeons. To rule out these effects, total MLCI values were compared among the members of each group, according to whether they acted as the surgeon in week 1, 2, 3 or 4, to ascertain whether there was a significant difference in their surgical proficiency, by using the one-way ANOVA with a significant value p < 0.05.

Once the effects of previous observation and/or assistance were ruled out, the MLCI values of each individual anatomical plane and the totals were compared between the control and the experimental groups, by means of the ANOVA test (significant value p < 0.05). All tests were performed by using the PAST® program.11

Ethical and animal welfare considerations

The present protocol was approved by the Comité Interno para el Cuidado y Uso de los Animales de Laboratorio (Internal Committee for the Welfare and Use of Laboratory Animals) from the Universidad Autónoma Metropolitana Unidad Xochimilco, with
reference number DCBS.CICUAL.02.10.

Results of simulator use

Comparisons of the proficiency of group members according to the week in which they acted the role of surgeon did not show a difference (p > 0.05), supporting the idea that observation and/or assistance did not improve technique. When comparing the MLCI values of each plane as well as total MLCI values between the control group and the experimental group, all were different (p < 0.05) with a higher rate of closure for the experimental group. The MLCI values of each group, as well as their comparisons, are shown in Table 1.

The experimental group performed the three planes of laparotomy suture in 5.34 ± 1.63 minutes per linear centimetre of incision (MLCI), compared to the control group that performed it in 7.03 ± 1.77 MLCI. This difference was significant (one-way ANOVA; p < 0.05) and showed that repeating the procedure three times with the simulator improved
suturing skills in a laparotomy.

 

Discussion

When comparing MLCI values among the participants of each group independently, and not presenting differences, it is evident that observing and/or helping during the procedure did not render psychomotor skills or abilities in the participants. The use of complementary strategies, such as the use of the simulator, is necessary for a student to develop manual dexterity and the instrument skills required for the successful application of sutures.1,6

On the other hand, the experimental group demonstrated better suture skills for the laparotomy in rabbits after performing three repetitions of the procedure in the simulator. These findings agree with those reported by Aggarwal,12 who found in his study that laparoscopic surgeons required two repetitions of a particular procedure in a simulator, in order to learn it. The simulator required them to hold the tissue, lift it up, place a clip, and then cut; for trainees, seven repetitions were required to learn to perform the same procedure. However, we have to take into consideration that this particular procedure would have a longer learning curve than performing a suture.

Conclusions

The results make evident the advantages of the use of simulators, when recommended as training devices for undergraduate students. However, these models should be considered as complementary tools in the teaching of surgical procedures, for they help in the acquisition of skills and abilities that lead to better performance in real patients, and eventually reduce the number of training events that require the use of live animals.13

More studies are required to determine the time and number of necessary repetitions in training with these bench simulators, in order to reach an adequate level of proficiency. Further work will also be needed to make the simulators more realistic, and to investigate ways in which to take maximum advantage of this training tool.

Author for correspondence:
Dr Juan J. Pérez-Rivero
Departamento de Producción Agrícola y Animal
Universidad Autónoma Metropolitana Unidad
Xochimilco
Calzada del Hueso 1100
Colonia Villa Quietud
Delegación Coyoacán 04960
Mexico City
Mexico
E-mail: jjperez1_1999@yahoo.com

References

1 Langebæk, R., Eika, B., Jensen, A.L., Tanggaard, L., Toft, N. & Berendt, M. (2012). Anxiety  in veterinary surgical students: A quantitative study. Journal of Veterinary Medical Education 39, 331–340.
2 Smeak, D.D. (2007). Teaching surgery to the veterinary novice: The Ohio State University experience. Journal of Veterinary Medical Education 34, 620–627.
3 Russell, W.M.S. & Burch, R.L. (1959). The Principles of Humane Experimental Technique, 238pp. London, UK: Methuen.
4 Martinsen, S. & Jukes, N. (2005). Towards a humane veterinary education. Journal of Veterinary Medical Education 32, 454–460.
5 Kumar, A.M., Murtaugh, R., Brown, D., Ballas, T., Clancy, E. & Patronek, G. (2001). Client donation program for acquiring dogs and cats to teach veterinary gross anatomy. Journal of Veterinary Medical Education 28, 73–77.
6 Valliyate, M., Robinson, N.G. & Goodman, J.R. (2012). Current concepts in simulation and other alternatives for veterinary education: A review. Veterinarni Medicina 57, 325–337.
7 Perez-Rivero, J.J. & Rendón-Franco, E. (2012). Experience of the use of table-top simulators as alternatives in the primary surgical training of veterinary undergraduate students. ATLA 40, P10–P11.
8 Schout, B.M.A., Hendrickx, A.J.M., Scheele, F., Bemel mans, B.L.H. & Scherpbier, A.J. (2010). Validation and implementation of surgical simulators: A critical
review of present, past, and future. Surgical Endoscopy 24, 536–546.
9 Griffon, D.J., Cronin, P., Kirby, B. & Cottrell, D.F. (2000). Evaluation of a hemostasis model for teaching ovariohysterectomy in veterinary surgery. Veterinary Surgery 29, 309–316.
10 Perez-Rivero, J.J. & Rendón-Franco, E. (2014). Cardiorespiratory evaluation of rabbits (Oryctolagus cuniculus) anesthetized with a combination of tramadol, acepromazine, xylazine and ketamin3. Archivos de Medicina Veterinaria 46, 145–149.
11 Hammer, Ø., Harper, D.A.T. & Ryan, P.D. (2001). PAST: Paleontological statistics software package for education and data analysis. Paleontología Electrónica 4, 1–9.
12 Aggarwal, R., Grantcharov, T.P., Eriksen, J.R., Blirup, D., Kristiansen, V.B., Funch-Jensen, P. & Darzi, A. (2006). An evidence-based virtual reality training program for novice laparoscopic surgeons. Annals of Surgery 244, 310–314.
13 Denadai, R., Oshiiwua, M. & Saad-Hossne, R. (2014). Teaching elliptical excision skills to novice medical students: A randomized controlled study comparing low- and high-fidelity bench models. Indian Journal of Dermatology 59, 169–175.

 

 

Rationalisation and Intellectualisation

Michael Balls

Russell and Burch saw failure to accept the correlation
between humanity and efficacy as an example of
rationalisation, a psychological defence mechanism

While wondering what I could discuss in this column I looked, as I often do, in the abridged version1 of The Principles of Humane Experimental Technique,2 at Russell and Burch’s introduction of what I call the humanity criterion. It is part of their discussion of the sociological factors which are among the Factors Governing Progress. This is how part of page 101 of the abridged version reads:
In fact, really informative experiments must be as humane as would be conceivable possible, for science and exploration are indissolubly linked to the social activity of cooperation, which will find its expression in relation to other animals, no less than to our fellow humans. Conscious good will and the social operational method are useless as safeguards against the mechanism of rationalisation (in the pathological sense of the term — i.e. the mechanism of defence by which unacceptable thoughts or actions are given acceptable reasons to justify them to oneself and to others, while, at the same tie, unwittingly hiding the true, but unconscious, motives for them).

The bold type indicates my explanation, and I have to admit that, six years after preparing the abridged version of The Principles, I now found it difficult to see what Russell and Burch had intended to convey. I therefore looked back at the original book, and found this paragraph on pages 156−157:
In efficacy, or yield of information, the advantages of humane technique apply almost universally. The correlation between humanity and efficacy has appeared so often in this book that we need not labour the point. There is, however, a more fundamental aspect of this correlation, specially important in research. Science means the operational method — telling somebody else how to see what you saw. This method is one of the greatest of all human evolutionary innovations. It has, however, one drawback. It prevents permanent acceptance of false information, but it does not prevent wastage of time and effort. The activity of science is the supreme expression of the human exploratory drive, and as such it is the subject to the same pathology. The scientist is liable, like all other individuals, to block his exploration on some front where his reactions to childhood social experiences are impinged upon. When this happens to the experimental biologist, we can predict the consequence with certainty. Instead of really exploring, he will, in his experiments, act out on his animals, in a more or less symbolic and exaggerated way, some kind of treatment which he once experienced in social intercourse with his parents. He can rationalise this as exploration, and hence fail to notice the block. But in fact such acting out invariably occurs precisely when real exploration is blocked, and must be relinquished before real exploration can begin again. Hence, such experiments will be utterly wasteful, misleading, and uninformative. The treatment of the animals, for one thing, will inevitably be such as to impair their use as satisfactory models. The interpretation of the results will be vitiated by projection. Really  informative experiments, must in fact be as humane as would be conceivably possible, for science and exploration are indissolubly linked to the social activity of cooperation, which will find its expression in relation to other animals no less than to our fellow humans. Conscious goodwill and the social operational method are useless as safeguards against the mechanism of rationalisation (in the pathological sense of the term).

Here, the underlining indicates what I omitted from the abridged version, and I now wonder why I did so. These words clearly reflect Russell’s interest in psychology — he later became a psychotherapist, and undoubtedly will have been influenced by discussions with his psychotherapist wife, Claire Russell. They could be seen as an explanation why some scientists did not appreciate the essential link between humanity and efficacy, and why Russell thought they needed what was offered by the Three Rs and the humanity criterion.

It is not clear what is meant by “the social operational method”, and consulting Google leads to only one hit — The Principles itself! “Conscious goodwill” is probably meant to contrast with unconscious rationalisation. Perhaps what Russell meant is that, however sincere the intention may appear to be, support for the Three Rs is useless, unless it leads to active and practical commitment to their development and application.

We are often confronted with rationalisation, the pseudo-rational justification of irrational acts,3 and its relative, intellectualisation, a different defence mechanism (or way of making excuses), “where reasoning is used to block confrontation with an unconscious conflict and its associated emotional stress, where thinking is used to avoid feeling. It involves removing one’s self, emotionally, from a stressful event. Intellectualisation is one of Freud’s original defence mechanisms. Freud believed that memories have both conscious and unconscious aspects, and that intellectualisation allows for the conscious analysis of an event in a way that does not provoke anxiety.”4

I am not a psychoanalyst, and I think it would be unwise, even dangerous, were I to seek to delve into the underlying reasons why some scientists are so keen to run to animal experimentation as the first resort and to do so little to make possible its replacement. Nevertheless, I can say, without fear of contradiction, that this is another great example of how Russell and Burch’s wonderful book continues to give us food for thought and calls for action.

Professor Michael Balls
E-mail: michael.balls@btopenworld.com

References
1 Balls, M. (2009). The Three Rs and the Humanity Criterion, 131pp. Nottingham, UK:  FRAME.
2 Russell, W.M.S. & Burch, R.L. (1959). The Principles of Humane Experimental Technique, xiv + 238pp. London, UK: Methuen.
3 Anon. (2015). Rationalization (psychology). San Francisco, CA, USA: Wikipedia Foundation, Inc. Available at: https://en.wikipedia.org/wiki/Rationalization_(psychology) (Accessed 26.08.15).
4 Anon. (2015). Intellectualization. San Francisco, CA, USA: Wikipedia Foundation, Inc. Available at: https://en.wikipedia.org/wiki/Intellectualization
(Accessed 26.08.15).
The Principles of Humane Experimental Technique is now out of print, but the full text can be found at http://altweb.jhsph.edu/pubs/books/humane_exp/het-toc. The abridged version, The Three Rs and the Humanity Criterion, can be obtained from
FRAME.

The choice of procedures

Michael Balls

The scientifically-justifiable choice of procedure
is a crucial issue in animal experimentation,
in the interests of both humanity and efficiency

In the opening of their chapter on Reduction and Strategy in Research in The Principles of  Humane Experimental Technique,1 Russell and Burch pointed out that “one general way in which great reduction can occur is by the right choice of strategies in the planning and performance of whole lines of research”.

They referred to the way Charles Hume had put it “in a searching essay”,2 that is, that “the central problem is that of choosing between trial and error on a grand scale and deductively inspired research”. The second type of choice can “take the form of testing deductions from well and consciously formulated hypotheses, or it may involve working from hunches — really the same thing, for where hunches are of any value, they are found to be based on equally precise hypotheses”. They said that the essence of the strategy “is that particular experiments are selected on some basis … from a larger set of experiments that  could have been performed”, and, as Hume had pointed out, “insighted” research “must be vastly less wasteful of animals”, where animals are to be used in the experiments.

I discussed this in an earlier comment on the Wisdom of Russell and Burch in relation to Reduction,3 with particular reference to experimental design and statistical analysis, but I now want to consider their discussion on The Choice of Procedures, in their chapter on Refinement.

Russell and Burch said that almost any research question “can always be answered in principle by a number of different procedures”, and the mark of distinction of “the great experimenter is the knack of choosing the most rapid, elegant and simple one”. But they then ask whether there are any simple rules in this context. One “general principle, important for both humanity and efficiency, is that of avoiding elaborate and roundabout methods”. Another rule is “the very careful formulation of questions”. One approach is to “first ask the question, then draw up, at least mentally, a list of procedures by which it could be answered”. If the list is long enough, consideration can be given to choosing the best procedure. If the list is too short, the question may need to be reformulated, to permit a wider range of procedural choice.

The greater the experimenter, in terms of ability and quality, the easier and better will be the choice of procedure. In addition, the humane experimenter will be careful to take account of Hume’s point about the wastage of animals, and Russell and Burch’s emphasis on the need for humanity and efficiency. This should be obvious, in terms of the use of resources, even if it were not a requirement of the laws under which animal experimentation is now permitted. It is a vital aspect of the education and training which must be regarded as essential for all those who are to undertake research, particularly if there is any risk of causing animal suffering, but also in the interests of the humans for whose benefit the research is being conducted.

It is encouraging that, in its response to the European Citizens’ Stop Vivisection Initiative,4,5 the European Commission has proposed four actions, one of which is to “analyse technologies, information sources and networks from all relevant sectors with potential impact on the advancement of the Three Rs”, in order to “present by end 2016, an assessment of options to enhance knowledge sharing among all relevant parties. The assessment will consider how  systematically to accelerate knowledge exchange through communication, dissemination, education and training.”

In order to aid researchers in their thought process when designing any project that could involve experimental animals, the FRAME Reduction Steering Committee designed a Strategic Planning Poster.6 The poster, which is available in several languages, guides the scientist through the decision points and steps needed when designing a whole programme of work, including the individual experiments within it. This resource, and the FRAME Training Schools in  Experimental Design and Statistical Analysis,7 encourage researchers to think about how to design their sequence of experiments, in order to minimise the number of animals that are exposed to the most severe procedures and to contemplate whether animals are needed at all.

Professor Michael Balls
c/o FRAME
Russell & Burch House
96–98 North Sherwood Street
Nottingham NG1 4EE
UK
E-mail: michael.balls@btopenworld.com

References
1 Russell, W.M.S. & Burch, R.L. (1959). The Principles of Humane Experimental Technique, xiv + 238pp. London, UK: Methuen.
2 Hume, C.W. (1957). The strategy and tactics of experimentation. The Lancet, 23 November, 1049–1052.
3 Balls, M. (2013). The Wisdom of Russell and Burch. 4. Reduction. ATLA 41, P24–P25.
4 Anon. (2015). Commission Response to the European Citizens’ Initiative “Stop Vivisection”. Brussels, Belgium: Intergroup on the Welfare and Conservation
of Animals. Available at: http://www.animalwelfare intergroup.eu/2015/06/05/european-commissiondoesnt- go-far-enough-to-meet-citizens-demands-toimprove-
animal-welfare/ (Accessed 08.07.15).
5 Balls, M. (2015). The European Citizens’ Stop Vivisection Initiative. ATLA 43, 147–150.
6 Gaines Das, R., Fry, D., Preziosi, R. & Hudson, M. (2009). Planning for reduction. ATLA 37, 27–32.
7 Fry, D., Gaines Das, R., Preziosi, R. & Hudson, M. (2010). Planning for refinement and reduction. ALTEX 27, Special Issue, 293–298.

 

In vitro models to mimic the endothelial barrier

Laurent Barbe, Mauro Alini, Sophie Verrier and Marietta Herrmann

Microfluidic technologies permit the replication in vitro
of geometrical features essential for the homeostasis of
all vascularised tissues in vivo, including the contribution
of pericytes to the endothelial barrier

Introduction

A functional microvasculature is critical for the homeostasis of all vascularised tissues.  accordingly, several diseases are associated with alterations in the microvasculature. For example, tumour angiogenesis is a major factor in determining the burden of the
disease. Furthermore, the formation of new vessels by angiogenesis and vasculogenesis is critical in the restoration of tissue function in ischaemic diseases. In tissue engineering, sufficient neovascularisation and early vessel anastomosis is thought to be a prerequisite
for the integration of the implant. These conditions have been extensively studied in animal models. However, in vivo studies have several limitations, including species differences and limited possibilities for imaging and tracking cells in the living animal. They also do not permit high-throughput and multiplexing applications. The development of microfluidic models of microvasculature and the endothelial barrier could help to overcome these problems and, most importantly, would replace a significant amount of animal experimentation. Nevertheless, microfluidic science is still an evolving research field, and many models do not address the endothelial barrier in its full complexity — for example, taking into account the contribution of pericytes.

The contribution of pericytes to the endothelial barrier

Pericytes are vascular mural cells associated with microvessels. The most common definition of pericytes goes back to their localisation, embedded in the endothelial basement membrane.1 Besides sharing the basement membrane, close interactions between endothelial cells and pericytes have been described, such as peg–socket contacts representing tight-, gap- and adherence-junctions.2 The relationship between the two cell types, particularly the anatomy of the pericyte coverage, reflects the function of the individual tissues. In organs with high exchange rates of gas and metabolites, the  distribution of pericytes is such that diffusion is minimally hindered.1, 3 Pericytes play an important role in vessel stabilisation, and underlying molecular signalling pathways have been described. Here, signalling through angiopoietin and Tie2 is critical; mutation of either angiopoietin 1 or Tie2 leads to mid-gestational death by cardiovascular failure in mice.2 In angiogenesis, the recruitment of pericytes is required for the stability of newly-formed vessels. Various studies have identified platelet-derived growth factor B (PDGF-B) as a major chemoattractant for pericytes. 4  Capillary and arteriolar pericytes also play an important role in inflammatory processes by guiding extravasating leukocytes toward their target. 5  Pericytes are also involved in several disease states, including airway remodelling in chronic allergic asthma6 and stroke.7 The identification of mesenchymal
stem cells and progenitor cells at perivascular sites, and the subsequent isolation and characterisation of such cells, suggests that pericytes have a role as multipotent progenitors.8-10

Current animal models

The function of pericytes in physiologically and pathologically relevant situations has been studied in transgenic mouse models containing LacZ-expressing or fluorophore-expressing pericytes.5, 6, 11 The migration and stimulation of pericytes in different disease situations was subsequently studied by the administration of pro-inflammatory factors, or by  backcrossing mice to a transgenic disease mouse model.5, 12  Interactions between labelled leukocytes and pericytes have been studied by lethal irradiation and the subsequent injection of bone-marrow cells from GFP mice.5 In addition, specific knockout lines have been used to study specific components of the cell junction and signalling complexes involved in pericyte– endothelial cell crosstalk, e.g. Sparc-deficient and Ccn2-deficient mice.13,14 Various animal models have also been developed, for studying the basic mechanisms of angiogenesis and vessel sprouting, including the cornea model, the chick chorioallantoic membrane (CAM) model, matrigel plug assays, and the dorsal skin fold chamber.15

Current microvascular models

Many factors determine the phenotypes of cells in vivo, including cell–cell interactions, interaction with the surrounding extracellular matrix, and the influence of various paracrine factors. Current cell culture vessels (dishes, flasks) cannot recapitulate these
aforementioned complex interactions. However, in recent years, microfluidic technologies have shown the potential to more-closely mimic the cellular microenvironment, at both the spatial and temporal levels.16 Typical microfluidic systems have geometrical features ranging in size from tens to hundreds of microns, and can host single cells or millions of cells arranged in a 2-D or a 3-D fashion. Microfluidic devices can accommodate flow control and therefore induce shear stress, which is known to have significant effects on the endothelial layer.17 This shear stress is absent in classical culture dishes. Furthermore, microfluidic technologies enable the generation of gradients over long periods of time, and also permit the control of paracrine factors in complex co-culture systems.18

In the past, various strategies were pursued to generate perfused microvessels in vitro.19 Different  levels of complexity were achieved, ranging from straight or branched channels within a material such as polydimethylsiloxane (PDMS), toward more complex microvascular networks within extracellular matrix hydrogels. In order to study cell–cell and cell– matrix interactions, microfluidic chips were designed that comprised two or more parallel channels, allowing the seeding of different cell types or hydrogels next to each other.20-22 Chen et al. reported on an alternative approach to studying the interactions between endothelial cells and other cell types.23  Here, each cell type was seeded in a different layer of the microfluidic device, separated by porous membranes; perfusion was applied to the endothelial layer of the chip.23 The embedding of microvessels in a 3-D matrix, potentially mimicking the perivascular tissue, represents another level of complexity. Several groups have addressed this challenge by generating endothelial cell-aligned microchannels within a collagen hydrogel.24–28 Such a set-up also allows the incorporation of pericytes in the hydrogel, which might eventually reassemble at the perivascular site of the endothelial cell layer.24, 26, 28, 29

Conclusions

Microfluidic technologies permit the replication of geometrical features found in vivo (i.e. small channels mimicking capillaries). In addition, due to the small dimensions involved, screenings are much less costly and time-consuming, as compared to similar in vivo studies. This is particularly interesting for applications such as testing the delivery of drugs. For the in vitro models to be successful, it is critical to consider all parameters of the endothelial barrier, including cell–cell and cell–matrix interactions, and the surrounding perivascular tissue.

Acknowledgements

The research of the authors is supported by the AO Foundation and the 3R Research Foundation Switz erland (Reduction, Refinement and Replacement of animal experimentation).

Dr Laurent Barbe
CSEM
Bahnhofstrasse 1
7302 Landquart
Switzerland

Dr Mauro Alini
AO Research Institute
Clavadelerstrasse 8
7270 Davos Platz
Switzerland

Dr Marietta Herrmann
AO Research Institute
Clavadelerstrasse 8
7270 Davos Platz
Switzerland

Author for correspondence:
Dr Sophie Verrier
AO Research Institute
Clavadelerstrasse 8
7270 Davos Platz
Switzerland
E-mail: sophie.verrier@aofoundation.org

References
1 Armulik, A., Genové, G. & Betsholtz, C. (2011). Pericytes: Developmental, physiological,  and pathological perspectives, problems, and promises. Developmental Cell 21, 193–215.
2 Armulik, A., Abramsson, A. & Betsholtz, C. (2005). Endothelial/pericyte interactions. Circulation Res earch 97, 512–523.
3 Allt, G. & Lawrenson, J.G. (2001). Pericytes: Cell biology and pathology. Cells, Tissues, Organs 169, 1–11. P36
4 Aguilera, K.Y. & Brekken, R.A. (2014). Recruitment and retention: Factors that affect pericyte migration. Cellular & Molecular Life Sciences 71, 299–309.
5 Stark, K., Eckart, A., Haidari, S., Tirniceriu, A., Lorenz, M., von Brühl, M.L., Gärtner, F., Khandoga, A.G., Legate, K.R., Pless, R., Hepper, I., Lauber, K., Walzog, B. & Massberg, S. (2013). Capillary and arteriolar pericytes attract innate leukocytes exiting through venules
and ‘instruct’ them with pattern-recognition and motility programs. Nature Immunology 14, 41–51.
6 Johnson, J.R., Folestad, E., Rowley, J.E., Noll, E.M., Walker, S.A., Lloyd, C.M., Rankin, S.M., Pietras, K., Erik sson, U. & Fuxe, J. (2015). Pericytes contribute to airway remodeling in a mouse model of chronic allergic asthma. American Journal of Physiology. Lung Cellular
& Molecular Physiology 308, L658–L671.
7 Bai, Y., Zhu, X., Chao, J., Zhang, Y., Qian, C., Li, P., Liu,  D., Han, B., Zhao, L., Zhang, J., Buch, S., Teng, G., Hu, G. & Yao, H. (2015). Pericytes contribute to the disruption of the cerebral endothelial barrier via increasing VEGF expression: Implications for stroke. PLoS One 10, e0124362.
8 da Silva, M.L., Caplan, A.I. & Nardi, N.B. (2008). In search of the in vivo identity of mesenchymal stem cells. Stem Cells 26, 2287–2299.
9 Crisan, M., Yap, S., Casteilla, L., Chen, C.W., Corselli, M., Park, T.S., Andriolo, G., Sun, B., Zheng, B., Zhang, L., Norotte, C., Teng, P.N., Traas, J., Schugar, R., Deasy, B.M., Badylak, S., Buhring, H.J., Giacobino, J.P., Lazzari, L., Huard, J. & Péault, B. (2008). A perivascular
origin for mesenchymal stem cells in multiple human organs. Cell Stem Cell 3, 301–313.
10 Chen, W.C., Park, T.S., Murray, I.R., Zimmerlin, L., Lazz ari, L., Huard, J. & Péault, B. (2013). Cellular kinetics of perivascular MSC precursors. Stem Cells International
2013, 983059.
11 Tidhar, A., Reichenstein, M., Cohen, D., Faerman, A., Copeland, N.G., Gilbert, D.J., Jenkins, N.A. & Shani, M. (2001). A novel transgenic marker for migrating limb muscle precursors and for vascular smooth muscle cells. Developmental Dynamics 220, 60–73.
12 Pfister, F., Feng, Y., vom Hagen, F., Hoffmann, S., Molema, G., Hillebrands, J.L., Shani, M., Deutsch, U. & Hammes, H.P. (2008). Pericyte migration: A novel mechanism of pericyte loss in experimental diabetic retinopathy. Diabetes 57, 2495–2502.
13 Rivera, L.B. & Brekken, R.A. (2011). SPARC promotes pericyte recruitment via inhibition of endoglin-dependent TGF-beta1 activity. Journal of Cell Biology 193, 1305–1319.
14 Hall-Glenn, F., De Young, R.A., Huang, B.L., van Handel, B., Hofmann, J.J., Chen, T.T., Choi, A., Ong, J.R., Benya, P.D., Mikkola, H., Iruela-Arispe, M.L. & Lyons, K.M. (2012). CCN2/connective tissue growth factor is essential for pericyte adhesion and endothelial basement membrane formation during angiogenesis. PLoS One 7, e30562.
15 Staton, C.A., Reed, M.W. & Brown, N.J. (2009). A critical analysis of current in vitro and in vivo angiogenesis assays. International Journal of Experimental Patholology 90, 195–221.
16 Bersini, S. & Moretti, M. (2015). 3D functional and perfusable microvascular networks for organotypic microfluidic models. Journal of Materials Science. Materials in Medicine 26, 5520.
17 Song, J.W. & Munn, L.L. (2011). Fluid forces control endothelial sprouting. Proceedings  of the National Academy of Sciences of the USA 108, 15,342–15,347. 18 Bae, H., Puranik, A.S., Gauvin, R., Edalat, F., Carrillo-Conde, B., Peppas, N.A. & Khademhosseini, A. (2012).
Building vascular networks. Science Translational Medicine 4,160ps123.
19 Zervantonakis, I.K., Kothapalli, C.R., Chung, S., Sudo, R. & Kamm, R.D. (2011). Microfluidic devices for studying heterotypic cell–cell interactions and tissue specimen cultures under controlled microenvironments. Biomicrofluidics 5, 13,406.
20 Bischel, L.L., Young, E.W., Mader, B.R. & Beebe, D.J. (2013). Tubeless microfluidic angiogenesis assay with three-dimensional endothelial-lined microvessels.
Biomaterials 34, 1471–1477.
21 Jeon, J.S., Zervantonakis, I.K., Chung, S., Kamm, R.D. & Charest, J.L. (2013). In vitro model of tumor cell extravasation. PLoS One 8, e56910.
22 Zervantonakis, I.K., Hughes-Alford, S.K., Charest, J.L., Condeelis, J.S., Gertler, F.B. & Kamm, R.D. (2012). Three-dimensional microfluidic model for tumor cell
intravasation and endothelial barrier function. Proceedings of the National Academy of Sciences of the USA 109, 13,515–13,520.
23 Chen, M.B., Srigunapalan, S., Wheeler, A.R. & Simmons, C.A. (2013). A 3D microfluidic platform incorporating methacrylated gelatin hydrogels to study physiological
cardiovascular cell–cell  interactions. Lab on a Chip 13, 2591–2598.
24Chrobak, K.M., Potter, D.R. & Tien, J. (2006). Formation of perfused, functional microvascular tubes in vitro. Microvascular Research 71, 185–196.
25 Price, G.M., Wong, K.H., Truslow, J.G., Leung, A.D., Acharya, C. & Tien, J. (2010). Effect of mechanical factors on the function of engineered human blood microvessels in microfluidic collagen gels. Biomaterials 31, 6182–6189.
26 van der Meer, A.D., Orlova, V.V., ten Dijke, P., van den Berg, A. & Mummery, C.L. (2013). Three-dimensional co-cultures of human endothelial cells and embryonic stem cell-derived pericytes inside a microfluidic device. Lab on a Chip 13, 3562–3568.
27 Morgan, J.P., Delnero, P.F., Zheng, Y., Verbridge, S.S., Chen, J., Craven, M., Choi, N.W., Diaz-Santana, A., Kermani, P., Hempstead, B., López, J.A., Corso, T.N., Fischbach, C. & Stroock, A.D. (2013). Formation of microvascular networks in vitro. Nature Protocols 8,
1820–1836.
28 Zheng, Y., Chen, J., Craven, M., Choi, N.W., Totorica, S., Diaz-Santana, A., Kermani, P., Hempstead, B., Fischbach- Teschl, C., López, J.A. & Stroock, A.D. (2012). In vitro
microvessels for the study  of angiogenesis and thrombosis. Proceedings of the National Academy of Sciences of the USA 109, 9342–9347.
29 Bichsel, C.A., Hall, S.R., Schmid, R.A., Guenat, O. & Geiser, T. (2015). Primary human lung pericytes support and stabilize in vitro perfusable microvessels. Tissue Engineering. Part A [E-pub ahead of print.]

Download a pdf of this article here.

Single housing of primates in US laboratories: a growing problem with shrinking transparency

Alka Chandna, Michael Niebo, Stacy Lopresti-Goodman and Justin Goodman

Thirty years ago, the United States took steps to enhance
the psychological well-being of primates in laboratories,
including the introduction of social housing requirements.
Now, in an apparent response to questions about
the effectiveness of these measures, federal authorities
are completely shutting down public access to
information on the implementation of social housing

Historical context

In 1985, in response to high-profile cases documenting the mistreatment of non-human primates used in experiments, 1 the US Congress amended the federal Animal Welfare Act (AWA) to mandate that institutions take steps to promote the psychological wellbeing of primates in laboratories.2 As a result, the US Department of Agriculture (USDA) created regulations pertaining to environment enhancement for primates, including provisions aimed at addressing their social needs.3 In federal regulations and guidelines, social housing of primates is now deemed to be the ‘default’, with exemptions permitted only for veterinary or experimental reasons with appropriate documentation and approval.4 In cases in which an experiment-related exemption was granted, institutions have been required to submit these written and approved justifications to the USDA with their annual reports on animal use. These documents — which report the number of primates singly housed for experimental reasons, and why it was deemed necessary — were then made publicly available.

Animal welfare science

The USDA’s primate social housing requirements are evidence-based, as social housing is universally acknowledged to be a key factor in the welfare of primates in laboratories. Moreover, it is well documented that housing primates alone, or single housing, is detrimental to their development, physical health, and psychological well-being. In rhesus macaques, physical contact with conspecifics is essential to normal development, and the amount of time spent caged alone is a significant predictor of stereotypic and self-injurious behaviour — including repeated pacing, circling, hyper-aggression, depression, hair plucking, or self-biting.5, 6  Psychological distress stemming from being caged alone has also been documented in cynomolgus monkeys,7  pigtailed macaques,8 chimpanzees,9 – 11 and baboons.12, 13 In one modified preference test involving capuchin monkeys, the value of social companionship was so high that the primates chose it in lieu of food.14 Singly housed primates have also been documented as having suffered from physiological abnormalities, including depressed immune function and higher incidence of coronary atherosclerosis.15, 16

Surveys on housing of non-human primates

Despite this evidence of the harms caused by single housing, as well as federal regulations and guidelines promoting social housing, the data indicate that many primates in US laboratories continue to be housed alone. A 2000–2001 USDA survey found that 34.7% of primates in US laboratories were housed individually — although the USDA admitted that this was likely a low estimate, as primates who had been housed only temporarily with other primates for breeding purposes were classified as being socially housed.17 A 2003 survey found that 54% of primates at 22 laboratories were singly housed — although this was also likely to have been a low estimate, as the study ill-advisedly included in its definition of “social housing” instances of “grooming-contact” housing, in which singly housed primates have some limited tactile contact with one another through barred or mesh barriers.18 A survey of primate housing at the Washington National Primate Research Center from 2004 to 2006 found that at least 63% of the monkeys were singly caged.19

An ongoing concern is that laboratories sometimes permit single housing of primates for the sake of convenience rather than necessity. For example, many laboratories singly house primates who have had surgical implants, such as head posts or other equipment, even though it is possible to successfully house them socially.20

Given the government mandate that social housing of primates in laboratories should be the default position, the aforementioned figures are cause for concern, particularly because they indicate that rates of single housing may be increasing.

Preliminary analysis of primate single-housing data

In the interest of conducting a current and more comprehensive evaluation of trends in primate single housing and the justifications provided, we attempted to undertake a new analysis of all single housing exceptions submitted in annual reports by laboratories to the USDA from 2010 to 2013 — the years for which data are currently available online. While the total number of facilities that confined primates (191 in 2010, 188 in 2011, 192 in 2012, and 184 in 2013) stayed relatively flat over the four years, there was a steady increase in the number of facilities that reported single-housing exceptions, from 30 (16%) in 2010 to 53 (29%) in 2013. However, when we attempted to look more closely at these exceptions, in order to determine trends in the numbers of singly housed primates and the  explanations given, we found glaring inadequacies in the data available on the USDA site. This inadequacy stemmed mainly from the failure of laboratories in meeting reporting requirements. From 2010 to 2013, the percentage of laboratories reporting singly
housed primates that failed to specify the number of primates singly housed for experimental reasons and the scientific justification for it — both required by law — increased from 36% to 47%. Worryingly, it also appeared that some, or all, of the required information was improperly redacted from many facilities’ reports. We also observed that several facilities that had provided very detailed exception letters in 2010 produced only very vague information in 2013 or, as noted above, redacted all the information, showing
a growing trend toward secrecy.

Government response

In December 2014, we informed the USDA of the problematic reports and requested the agency’s assistance  in securing the missing data. In February 2015, we received correspondence from the USDA (a personal communication), which stated, “We have had
discussions and are finalising our position.” We heard nothing further, but in March, the USDA sent an email to all its stakeholders announcing changes to its Inspection Guide — used by its inspectors to identify violations of the AWA at the facilities they inspect.
The announcement noted that the revisions included a “more consistent procedure for reporting exceptions and exemptions on the Annual Report.”21 Upon closer examination, it became clear that the revised inspection guide now specified that exemption of a primate from “some or all of the environmental enhancement plan” — which would include social housing — “should not be reported on the Annual Report” [emphasis in the original].22 Prior to the recent revision to the inspection guide, it was standard operating procedure for laboratories to report single housing of primates.

Rather than ensuring that laboratories were properly reporting single housing of primates, the USDA instead — we suspect in consultation with the laboratory community — took the backward step of simply exempting facilities from having to submit this information at all.

Discussion

A 2011 study seeking to assess the effectiveness of the 1985 AWA primate psychological health amendments determined that “the current system of laboratory animal care and record keeping is inadequate to properly assess AWA impacts on primate psychological wellbeing and that more is required to ensure the psychological well-being of primates.”17 It was already difficult to ascertain this information, and now the USDA has made this task virtually impossible. Even with the reporting requirement in place, laboratories were often failing to approve, document, and report the single housing of primates. For instance, a 2011 USDA inspection report for a US contract laboratory cited the company for singly housing 83% of the more than 6,000 primates at the facility  without securing the necessary justifications or reporting the matter to the USDA.23,24 This number amounts to more than 4% of all primates housed in US laboratories.

Any previously reported figures are also underestimations, because institutions have only been required to report the numbers of primates who are singly housed for ‘experimental’, but not veterinary, reasons; this could account for a third more singly housed primates.25

Conclusions

On the 30th anniversary of the amendments to promote the psychological well-being of primates, the evidence that singly housed primates suffer is overwhelming, as is the proof that the US government and laboratories are failing to confront this rapidly growing problem effectively. The USDA’s recent revisions to reporting requirements will limit the availability of data, and consequently will stifle informed debate on the suffering of primates in laboratories and failures of the existing regulatory system. To address this issue meaningfully, we need more transparency and accountability, not less.

Author for correspondence:
Dr Alka Chandna
People for the Ethical Treatment of Animals
501 Front Street
Norfolk, VA 23510
USA
E-mail: AlkaC@peta.org

Michael Niebo
People for the Ethical Treatment of Animals
501 Front Street
Norfolk, VA 23510
USA

Dr Stacy Lopresti-Goodman
Department of Psychology
Marymount University
2807 N. Glebe Road
Arlington, VA 22207
USA

Justin Goodman
People for the Ethical Treatment of Animals
501 Front Street
Norfolk, VA 23510
USA
and
Department of Sociology
Marymount University
2807 N. Glebe Road
Arlington, VA 22207
USA

References

1 Carbone, L. (2004). What Animals Want: Expertise and Advocacy in Laboratory Animal Welfare Policy, 304pp. Oxford, UK: Oxford University Press.
2 US Congress (1985). Public Law 99-198, Food Security Act of 1985, Subtitle F — Animal Welfare. Title XVII. Available at: https://awic.nal.usda.gov/public-law-99-198-food-security-act-1985-subtitle-f-animalwelfare (Accessed 31.05.15).
3 US Department of Agriculture (1991). Final Rules: Animal Welfare; Title 9, CFR (Code of Federal Register) Part 3. Standards. Federal Register 55 (No. 32), 6426–6505. Available at: https://awic.nal.usda. gov/final-rules-animal-welfare-9-cfr-part-3 (Accessed 31.05.15).
4 National Research Council (2011). Guide for the Care and Use of Laboratory Animals: Eighth Edition, 248pp. Washington, DC, USA: The National Academies Press.
5 Baker, K.C., Bloomsmith, M.A., Oettinger, B., Neu, K., Griffis, C., Schoof, V. & Maloney, M. (2012). Benefits of pair housing are consistent across a diverse population of rhesus macaques. Applied Animal Behaviour Science 137, 148–156.
6 Bayne, K. (2005). Macaques. From the booklet series Enrichment for Nonhuman Primates. Washington, DC, USA: Department of Health and Human Services. Available at: http://grants.nih.gov/grants/olaw/ Enrichment_for_Nonhuman_Primates.pdf (Accessed 05.04.15).
7 Shively, C.A., Clarkson, T.B. & Kaplan, J.R. (1989). Social deprivation and coronary artery atherosclerosis in female cynomolgus monkeys. Atherosclerosis 77, 69–76.
8 Bellanca, R.U. & Crockett, C.M. (2002). Factors predicting increased incidence of abnormal behavior in male pigtailed macaques. American Journal of Primatology 58, 57–69.
9 Brent, L., Lee, D.R. & Eichberg, J.W. (1989). The effects of single caging on chimpanzee behavior. Laboratory Animal Science 39, 345–346.
10 Kalcher, E., Franz, C., Crailsheim, K. & Preuschoft, S. (2008). Differential onset of infantile deprivation produces distinctive long‐term effects in adult ex‐laboratory chimpanzees (Pan troglodytes). Developmental Psychobiology 50, 777–788.
11 Nash, L.T., Fritz, J., Alford, P.A. & Brent, L. (1999). Variables influencing the origins of diverse abnormal behaviors in a large sample of captive chimpanzees (Pan troglodytes). American Journal of Primatology 48, 15–29.
12 Coelho, A.M., Carey, K.D., & Shade, R.E. (1991). Assessing the effects of social environment on blood pressure and heart rates of baboons. American Journal of Primatology 23, 257–267.
13 Kessel, A. & Brent, L. (2001). The rehabilitation of captive baboons. Journal of Medical Primatology 30, 71–80.
14 Dettmer, E. & Fragaszy, D. (2000). Determining the value of social companionship to captive tufted capuchin monkeys (Cebus apella). Journal of Applied Animal Welfare Science 3, 293–304.
15 Lilly, A.A., Mehlman, P.T. & Higley, J.D. (1999). Traitlike immunological and hematological measures in female rhesus across varied environmental conditions. American Journal of Primatology 48, 197–223.
16 Doyle, L.A., Baker, K.C. & Cox, L.D. (2008). Physiological and behavioral effects of social introduction on adult male rhesus macaques. American Journal of Primatology 70, 542–550.
17 Balcombe, J., Ferdowsian, H. & Durham, D. (2011). Self-harm in laboratory-housed primates: Where is the evidence that the Animal Welfare Act amendment has worked? Journal of Applied Animal Welfare Science 14, 361–370.
18 Baker, K.C., Weed, J.L., Crockett, C.M. & Bloomsmith, M.A. (2007). Survey of environmental enhancement programs for laboratory primates. American Journal of Primatology 69, 377–394.
19 Thom, J.P. & Crockett, C.M. (2008). Managing environmental enhancement plans for individual research projects at a national primate research center. Journal of the American Association for Laboratory Animal Science 47, 51.
20 DiVincenti, L., Jr & Wyatt, J.D. (2011). Pair housing of macaques in research facilities: A science-based review of benefits and risks. Journal of the American Association for Laboratory Animal Science 50, 856.
21 USDA (2015). USDA Revises Its Inspection Guide to Improve Oversight of Research Facilities. Washington, DC, USA: United States Department of Agriculture. Available at: http://content.govdelivery.com/ accounts/USDAAPHIS/bulletins/f4e94e (Accessed 31. 05.15).
22 USDA (2015). Animal Welfare Inspection Guide, 424pp. Washington, DC, USA: United States Department of Agriculture. Available at: http://www.aphis. uda.gov/animal_welfare/downloads/Animal%20Care %20Inspection%20Guide.pdf (Accessed 31.05.15).
23 USDA (2011). Inspection Report for SNBL USA, Ltd. Washington, DC, USA: United States Department of Agriculture. Available at: http://www.mediapeta. com/peta/PDF/July132011-78percent singly housed.pdf (Accessed 25.06.15).
24 USDA (2011). Annual Report for SNBL USA, Ltd. Washington, DC, USA: United States Department of Agriculture. Available at: http://www.mediapeta. com/peta/PDF/SNBL-AnnualReportfor2011.pdf
(Accessed 25.06.15).
25 Lee, D.R. (2013). Social housing strategies for nonhuman primates [PowerPoint slides]. Available at: http://www.aclam.org/content/files/files/forum2013/aclam_forum_2013_lee_social.pdf (Accessed 05.04. 15).

Download a pdf of the article here: Primate single housing.