Point/Counterpoint: The Cases For and Against Harmonization

Publication Details

William S. Stokes


Michael D. Kastello

Gilles Demers

Jonathan Richmond

John G. Miller

Nelson Garnett

Wim deLeeuw

Derek Forbes

Clément Gauthier

John C. Crabbe

Naoko Kagiyama

DR. STOKES (William S. Stokes, National Institute of Environmental Health Sciences, Research Triangle Park, NC): I want to thank all who have stayed throughout the meeting to this final session, which I think will be well worth your attendance. We will discuss the cases for and against harmonization of laboratory animal guidelines and how we might approach balancing the need for accomplishing high-quality medical research while providing for optimal animal welfare.

First, I would like to comment that I have been very impressed with this conference and the level of the discussions that have been taking place. I think this workshop is moving us in the right direction. I have been involved in international harmonization for safety testing guidelines for the past 10 years. It is not an easy process, but when harmonized guidelines are finally established, it is a good feeling. So if we move in that direction in the area of animal care and use guidelines, I believe it will be, again, potentially challenging, but a rewarding and worthwhile experience.

Second, I am also impressed by the fact that there are nearly 150 people here, from 14 different countries. This attendance truly reflects an international commitment to animal welfare. Clearly, harmonization has the potential to contribute to increased replicability and comparability of research and testing results based on the use of animals. It also enhances the likelihood that there is a benchmark of animal care that is adhered to on an international basis.

In a perfect world, we would all sit down at the table, look at the scientific data, and reach a consensus on good animal care guidelines, based on that science. But as discussed at this meeting, our guidelines have evolved largely based on professional judgment and the ability and need to obtain animals that are healthy and free of injury, and to facilitate high-quality scientific research and testing. However, this workshop has identified data gaps that need to be filled to have a basis not only for future guidelines, but even for our current guidelines. Some issues have been raised about proposing changes, such as increased cost and the potential to introduce confounding factors or variables in our research and testing. For this reason, I believe that we must proceed with caution in harmonizing guidelines.

Nevertheless, we are at a critical crossroads as to how we move forward. You all have probably heard the famous saying of Yogi Berra: “When you come to a fork in the road, take it.” Clearly, there is a fork in the road, and each of us is going to take it. However, we need to proceed on an informed basis and be certain that the direction we choose is the correct one.

Prior to this meeting, we asked the panelists to describe their proposed position with regard to harmonization—whether they were for it or against it. We were hoping to have about six for and six against. As you can imagine, being opposed to harmonization does not sound very favorable. It is like not supporting “motherhood and apple pie.” In our results, many responders were for harmonization; some were for harmonization with reservations; and at least one responder was against harmonization. Based on these results, I think the discussion today will be lively.

We will use the following format: I will ask each of the panelists to state his/her position in 2 to 3 minutes, and then to comment on how s/he would advance that position (whether it is for harmonization, for harmonization with reservations, or against harmonization). Next, the panelists will be asked to respond to each other's comments, with the opportunity for questions. We then will take questions and comments from the audience. In this session, we will proceed with the participants in the order in which they are listed on the program.

DR. KASTELLO (Michael D. Kastello, Aventis Pharmaceuticals, Bridgewater, NJ): I would like to commend and thank ILAR for organizing this meeting. It has been very interesting and useful to me. I would consider it, in many ways, a first step in a dialogue that I think will persist for a long time.

I am responsible for research animals and animal care and use programs in France, Germany, Japan, and the United States. Harmonization, whatever it might be, is something that seems as if it should be useful to me, with the diverse responsibilities that I have. But this afternoon I am speaking against harmonization.

I do not know the meaning of “harmonization.” Is it animal care? Is it animal welfare? Is it laws, regulations, standards? Is it the minimum standard? Is it best practice?

With more than 30 years of experience, I continue to be amazed by what we have learned from animals about biological mechanisms, human and animal diseases, safety and efficacy of medicines and vaccines, and the risks of natural and man-made products. I am also impressed by what we do not know about the animals we use.

It has been stated repeatedly here the last few days that data gaps exist. I think that is a kind phrase. We have huge areas for which we have insufficient information. Some individuals in the audience have interpreted that statement as being a reason for not doing something; however, I think it is a reason for doing something.

More importantly, perhaps, than the data gaps are the knowledge gaps. We can ask, is one study enough, or do we need 50? Yet the application of that information is critical. The knowledge gaps are exacerbated, in my mind, by the cultural gaps, traditions, values, laws, and regulations that already exist, as well as religious differences.

This concept of harmonization is seductive, but we have much to do first. In my opinion, standards, guidelines, regulations, and laws must be science based. They must benefit the animals we use. They must not burden research unnecessarily. It is not sufficient to show that a practice we propose is harmful; I think we must show that it is beneficial to the animals.

We can never address all of the issues, conditions, circumstances, species, strains, and so forth. Therefore, whatever to do in moving forward must include the application of performance standards and professional judgment. I am not yet ready for harmonization.

DR. DEMERS (Gilles Demers, International Council for Laboratory Animal Science (ICLAS), Quebec, Canada): My comments are provided on behalf of the ICLAS. I will speak on behalf of all people around the world who are doing research. Some of them come from countries that are not well developed. They are using animals, and they are trying to find ways to do good science. For this reason, they have requested that ICLAS participate in the process of harmonization of guidelines.

The participation of ICLAS is important for them because their countries are different from ours. All of us have come here from countries that have legislation and guidelines in place, yet other countries not represented here have no guidelines and, in some cases, no legislation. They therefore want to follow what is done in Europe, or in America. However, they need documents to empower what they want to do, because they are working for universities where sometimes the concept of triage is not well supported. They seek statements that have been issued by an international scientific community.

In this workshop, we represent that type of community. We are a large group of individuals involved in several scientific organizations. We have the capacity and the power to make statements and to support documents and guidelines. Because they do not constitute legislation, it is possible for us to agree or disagree on a paragraph or a portion of a guideline. Basically, the guidelines exist to help people understand the ethical value of what they are doing. In that sense, I believe that we should express our support by working, through meetings like this one, for harmonization. Nevertheless, ICLAS opposes standardization because ICLAS is in constant liaison with countries and regions that have different cultures, traditions, religion, legislation, and regulations.

A recent example of these differences has been evident during this workshop in discussions with Dr. Kagiyama, who is from Asia and who will speak later. We have spent much time during these 3 days discussing environmental enrichment, although this concept does not really exist in the perspectives of Asian people. Asian scientists are focusing more on good science and the good quality of their animals' health. They want the biological material they use to be perfect. They understand that animals in good health are animals that are free of stress. To avoid stress, they must help the animals deal with their psychological and physiological needs. In this way, they are providing environmental enrichment. They are providing contact bedding, social housing, and visual access to other animals. They may not provide as many devices as are provided in other countries, but they are performing good science.

For these reasons, I believe that we should expand this recent dialogue in the future to include other countries. We must take into consideration that other countries and other regions of the world do not use the same words as we are using. But we are all in favor of harmonization.

DR. RICHMOND (Jonathan Richmond, Home Office, Dundee, United Kingdom): As a regulator, my position on harmonization is, “yes, but,” and the “buts” are as important as the “yes.” I think harmonization is defensible only if the objective is to promote the best science and the best animal welfare, not simply to standardize around something. When I say best animal welfare, I mean that which makes the animals feel good, not that which makes a regulator or an IACUC feel good.

I believe we have established that there is not enough evidence for all of these things to be evidence based. But I think in the absence of evidence, they have to at least be rational, defensible, and based on systems that have been tried and are known to work in practice. I think they should always, if possible, be in the form of performance, rather than engineering, standards. However, I think that on many fronts, we are not even ready for the standard. We are probably ready to agree only on the principle around which the standard should be stenciled. If there are performance standards, then I believe there must be clarity about how compliance will be determined and about the audit trail that will have to be maintained to be sure that people have made appropriate informed decisions where discretion is allowed.

I would take exception to a comment that has been made two or three times here. I do not believe there should be as much flexibility as possible. I believe there should be as much flexibility as is necessary to achieve two outcomes: first, to make sure that the arrangements that are made are the best context-specific arrangements required to promote welfare and good science and, second, to allow innovation so that we can continuously challenge and increase standards. But if you want to be an innovator, I think you also must be aware that the onus is on you to demonstrate the impact of your innovation, rather than simply to blaze a trail and hope others will follow.

We have also heard there are many things that can influence a data set, not all of which are recognized or captured at the moment. I will take you back again to the Russell and Burch book of 1959. They recognized the same thing. One of the things they pointed out about the Methods section of the publication is that your duty as an author is to use the Methods section not simply to describe what you did, but to provide enough detail so that other scientists can do what you did and see what you saw. You should provide ample information if you want others to be able to reproduce your work.

Let me leave you with this final thought, from someone who is a regulator. There are many gaps in the science behind the guidelines. In the real world, science, at best, informs rather than determines the outputs and the political process. But let me tell you this: In the absence of science and evidence, it is those with influence, not those with knowledge, who determine the output of that process.

DR. MILLER (John G. Miller, American Association for Accreditation of Laboratory Animal Care (AAALAC), International, Rockville, MD): As the Executive Director of AAALAC International, my position regarding harmonization is somewhat indescribable. Prior to this meeting, the discussions about harmonization of which I was aware focused on harmonization of standards. In that context, I think my “bottom line” is that there should be no mutually exclusive provisions, such as never allowing animals to be single housed.

We have already heard some of the other factors that must be considered in any discussion of harmonization, such as cultural and socio-economic differences. It is difficult to convince individuals and organizations in developing countries that animals should live better than a large percentage of their population. This difficulty might lead one to the least-common-denominator approach, but is that really where we want to be? We just heard from Dr. Richmond that he does not endorse that approach, which leads me to “Miller's law.” Miller's law states that the level of specificity in a standard, or indeed in any written document, is inversely proportional to the number of different viewpoints involved in the preparation. In other words, complete agreement results in lack of substance.

The exception to the Miller's law equation are the principles of the Committee of International Organizations of Medical Science (CIOMS) (CIOMS 1985). How many here know what the CIOMS principles are? As I suspected, very few of you are indicating that you know. They are excellent, and already agreed upon. CIOMS is part of the World Health Organization, so the United Nations and many countries are involved, and their principles are as specific as a multinational agreement can ever be. My position is that the likelihood of formulating meaningful harmonized written standards is extremely low, and even if it were possible in the real world, as Dr. Richmond points out, their application and implementation would undoubtedly be uneven, at best.

So what might be possible? I think it is possible to have harmonized practices. We already have examples of these practices—ethical review, animal care and use review, and national oversight authority—which can be harmonized and upheld by a recognized national authority. The practices can be based on harmonized principles; the CIOMS principles could be rejuvenated. An international organization such as AAALAC would be involved to assess and verify that the practices are, in fact, harmonized to a single standard that is not written, that is evolutionary, and that can change as we learn more about the area of animal care and use.

DR. GARNETT (Nelson L. Garnett, Office of Laboratory Animal Welfare (OLAW), National Institutes of Health (NIH), Bethesda, MD): It is extremely important to recognize that the mission of OLAW is animal welfare in the context of the mission of NIH, which is biomedical research. Partly for that reason, I believe that harmonization is extremely important for both scientific and welfare reasons, if only, for example, to improve the quality, reduce the variability, and increase the ability to compare results across countries. With regard to welfare, in addition to the potential benefits to the animals themselves, we have the issue of credibility and the integrity of the process for which we are striving to gain public support. Properly applied, harmonization could enhance all of those various things.

I do think it is very important that we have some recognition across jurisdictions about equivalency of several systems or international standards. Actually NIH already does that. Although most of the NIH funding is within the United States, there are nondomestic institutions that receive NIH funding, either directly or as performance sites. The recognition of the sovereignty and, essentially, of the standards of the performance-site country is built into our system. That recognition is firm, and it requires people to follow their own rules.

I agree that it is important to avoid engineering-based mandates that have numerous implications, both intended and unintended. I am thinking in particular of things like trade barriers, wherein artificial restrictions or limitations on the use or the sale of a product, for example, might be limited. I believe we need to avoid such examples.

I agree that we need science-based standards to the extent possible, but it is still extremely important to do the science. As we have heard over the last 3 days, interpreting what the science actually means requires tremendous care. What is it telling us? Are we looking at the results objectively? Perhaps more importantly, are we conscious of the fact that some self-interest may be involved? We must be careful to separate ourselves from conflict of interest in how we interpret and apply the knowledge we gain. Finally, I believe it is extremely important that we as a community identify the questions that need to be answered to inform public policy in this area.

DR. DELEEUW (Wim deLeeuw, Directorate of Inspection, Food & Consumer Safety Authority, Department for Veterinary Public Health, Animal Diseases, Animal Welfare & Feed, Den Haag, The Netherlands): As a public health inspector in the Netherlands and Chairman of the working party of the Council of Europe, I also am in favor of harmonization. I will not repeat the arguments already mentioned; however, I suggest that we might question the kind of harmonization we are discussing. First, in my opinion, it is not necessary to strive toward a harmonization of all of the regulatory instruments because, indeed, cultural and international differences play an important role. An attempt to force all of these differences into one process would not be appropriate and would be a very long process. Instead, I believe that harmonization should target starting points—criteria and opinions. Especially in the preparatory phase of such a process, this approach will lead to the exchange of opinions and thoughts, as we have experienced these last few days. That process in itself will lead to improvements of quality. If the process goes well, then we can take the best from several worlds. Whether the laboratory is in Berlin or in Washington, I believe there is a common goal—to define and provide for the needs of the laboratory animals. To that end, our common goal requires common, shared knowledge.

For me, and perhaps for the majority here, the benefits of these meetings have been quite variable. We had a very good exchange of information about the regulatory processes in our countries and on our continents, about ongoing revision processes, and about results of relevant research. Yet at the same time, it has been evident at this meeting that many differences still exist with regard to opinions, knowledge, and perception. The question is how we can progress in striving for harmonization. One idea is to reduce the attendance at future meetings in an effort to create a kind of intercontinental or transatlantic steering committee. A smaller-scale group could be responsible for staying on course, for taking initiatives without having to do all of the work themselves, with the aim of making their goal more visible. In this meeting, I have perceived a significant lack of knowledge in some areas, whereas in other areas the amount of knowledge appeared to be great, but either it is not available everywhere or it is not accepted to the same degree everywhere. So perhaps a smaller group could take the initiative to strive toward reaching agreement on research criteria for best practices, which could form the basis of harmonization, at least in opinions.

DR. FORBES (Derek Forbes, Division of Biomedical Services, University of Leicester, Leicester, United Kingdom): We have heard much discussion this week about the gaps we need to fill in science-based data. I believe some of those gaps are created by a lack of documentation of the science that has already been done. With regard to safety testing, I would take the Organisation for Economic Co-operation and Development (OECD) as an example of international regulations that are based on good laboratory practice. The essence of good laboratory practice is its documentation, which can be questioned and inspected. Different countries reciprocate each other's inspectors. The quality of the safety testing is paramount because the products derived from those safety tests and efficacy tests are applied internationally.

In terms of harmonization, I suggest that those of us who use animals, at whatever level of research, should have the benefit of sufficiently well-documented studies to respond to any type of interrogation. The scientific quality of the studies should be paramount. In addition, the scientific journals should assist us by publishing sufficient information for the reader to be able to judge the quality of the science and interpret it. This type of cooperation among scientists would close many of the gaps we have discussed.

DR. GAUTHIER (Clément Gauthier, Canadian Council on Animal Care (CCAC), Ottawa, Canada): I would like to congratulate ILAR and the National Academy of Sciences for having held this conference, which has been extremely useful. On behalf of CCAC, I would like to express qualified support for harmonization. By harmonization, we really mean harmonization of national standards, of guidelines, and of a concerted approach to reach common objectives, rather than “international standardization.” I agree with what Dr. Miller said about non-mutually exclusive positions. I believe that these guidelines should have the following three characteristics: They should (1) provide clear benefits to the animals, (2) not interfere with research, and (3) be based on science. But here I would like to depart slightly from my predecessors' statements. In addition to the publications written by authors who represent the conventional scientific disciplines (e.g., behavioral sciences, laboratory animal sciences) and all of the different types of knowledge that are not necessarily published in the mainstream scientific journals, I believe we should also recognize the side of scientific information that is coming and needed. Guidelines should be published and used as reference tools.

I recognize that some people might interpret the published guidelines as regulations, but I believe that this situation is a matter of perceptual distortion, which takes place at the individual level. I have faced that perceptual distortion in Canada several times, with individuals, for example, who are less qualified to be on an animal care committee than other members. Rather than understanding the basis and the rationale for CCAC guidelines, and because they often have not been involved early enough in the process, they try to place the onus on the CCAC. Regulatory agencies also face this situation, which I believe is a matter of perception. I agree with Dr. Richmond that in this context the people are that key, and their professional judgment makes the difference. This was the situation 30 years ago, and it is still the case, that we have insufficient collective knowledge to be able to draft, write, and agree on specific guidelines or on standards. We defer to professional judgment at the local level, and the substantive ethical decisions are made at the local level. Even if the scientific community is ready to progress collectively and worldwide, it cannot progress beyond the local level to another stage in the guidelines.

In conclusion, I would like to make the following three points: (1) We need to maintain a sensible basic approach throughout institutional animal care and use committees to develop and implement guidelines that incorporate the input of our peers. At the outset—not 10 years after the guidelines have been developed—we should include scientists, welfare organizations, and the public. (2) We need to exchange good practices across animal care committees and through international networks. Animal users in all countries should increase their understanding of their own national regulatory/oversight system. I have perceived a major deficiency in this context during these last 2 days of debate. If one lacks a basic understanding of how those guidelines are developed and when they should be applied, one cannot apply them or know their range of flexibility. Guidelines have flexibility in European countries. For example, in Europe I gained an understanding of the ban on the ascites method, which is equivalent to what we have in our recently developed guidelines. (3) We should increase the communication between regulatory administrators across countries on an ongoing basis. Dr. de Leeuw made this proposal, which Drs. Miller and Garnett also addressed. One simple way to accomplish that goal is to invite each other to hands-on experiences at the local level. For example, Dr. Garnett participated in one of our assessments at a large institution in Canada about 2 years ago, after which he basically confirmed that CCAC accreditation would serve the purposes of the Public Health Service.

DR. CRABBE (John C. Crabbe, Oregon Health and Science University and the Portland Veteran Affairs Medical Center, Portland, OR): I volunteered to speak against harmonization. Because you may characterize my position as essentially a straw-man argument, I will clarify it. I believe that the drive toward harmonization will inevitably lead to standardization. My concern is that standardization will further restrict the environmental range in which we study organisms that serve as our experimental subjects.

My particular interest is in understanding how genes influence behavioral endpoints, which leads inevitably to misunderstanding completely how genes, in fact, influence behavior, and under which circumstances. It can easily lead to false-positive findings, wherein one concludes that a gene affects a particular behavior. However, if everyone is using the same cage size and the same lack of enrichment, we are really studying the effects of those genes under a very small window of the total environment. It can also lead to false-negative findings, wherein one fails to detect true influences of either the biological manipulation in use or the genetic manipulation in use, because you looked in only one place, and the gene acts in many other places. Generally, because genes and biology interact with the environment of the animal, we need to explore environments and the biology of the organism systematically.

DR. KAGIYAMA (Naoko Kagiyama, Central Institute for Experimental Animals, Kawasaki, Japan): I would like to take a moderately against stance. I wonder to what extent we can harmonize the guidelines. Fortunately, the 3Rs principle has been adopted worldwide, so we may easily harmonize that principle in the guidelines. But what about the process of realizing the principle? Do you think we can harmonize the process to reach the goal? I daresay it is not easy. Each country has its own legal requirements, culture, and attitudes, which extend beyond science, are sometimes political, and are sometimes personal.

Actually, both the European countries and the United States are guiding the process independently by the ILAR Guide for the Care and Use of Laboratory Animals (Guide). ILAR has chosen a science-based performance approach. In reference to the approach in the Guide, I would like to mention and paraphrase an interesting sentence that appears on page 25: It says that animal performance indexes, such as health, reproduction, growth, behavior, activity, and use of space, can be used to assess the adequacy of housing. The Guide thus advises us to observe animals precisely to know how to treat the animals. I understand this to be the basic concept of a performance approach, and I believe that we might harmonize a process if we focus on a performance approach. In that case, dissemination of the latest knowledge and technology, as well as the information about the variety of regulations, would be essential for the harmonized guidelines. For these reasons, I am moderately against harmonization.

DR. STOKES: Thank you, Dr. Kagiyama. At this point, I would like to return to Dr. Miller, who wanted to add a few clarifying remarks about AAALAC and to pose a few questions to the rest of the panel.

DR. MILLER: I am a firm believer in the old marketing adage, that “it is bad publicity only when they misspell your name.” From that standpoint, I appreciate the frequent references to AAALAC during this workshop. Although I think that AAALAC's record, utility, and worth speak for themselves, I also believe there are some misconceptions, which I would like to correct for international colleagues who are less familiar with AAALAC that my US colleagues, and who may benefit for a brief review of AAALAC's role.

To begin with the basic system by which science is funded in the United States, people and companies pay taxes. The government decides how to allocate those funds, and it gives multiple billions of dollars every year to NIH, which then funds individual researchers and institutions. As in most instances, when you take public money, “strings” are almost always attached. In spite of the fact that we have a very decentralized system, some strings are attached. When you accept federal money to do research, you agree and promise, through your written assurance to Dr. Garnett's office, that you will follow the provisions of the Guide. You are not able to pick and choose which provisions you will follow and which ones you will ignore. You commit to follow them; therefore you must follow them. However, we have a system that specifically allows deviations from the provisions in the Guide, as long as they are scientifically justified and approved by the institutional animal care and use committee (IACUC). So there is a mechanism in place to deviate from what you have agreed to follow.

It is also very important for all of us to be reminded—because we tend to forget—that both the Public Health Service policy and the Animal Welfare Act put science above other requirements. Science will always prevail. The intent of both is to never interfere with science.

The role of AAALAC in this context is to assess the institution's adherence to agreed-upon standards. As mentioned, you have voluntarily invited visits from Dr. Garnett's office for their provision of an independent assessment. We therefore hold you to that standard, which has flexibility.

Finally, the Guide we use does not require transparent cages. It requires that cages allow observation of the animals with minimal disturbance to them. So if there is a way to accomplish that provision without cages being transparent, then you are free to act accordingly. I would encourage you to get your copy of the Guide and refer to page 36 to read about environmental enrichment. It is extremely flexible, as are most sections of the Guide, and it is open to a great deal of interpretation at your institution.

With regard to harmonization, I believe that many people are undecided at this point, and very few people enthusiastically support the concept. For those who are in favor of harmonization, I have the following series of queries, organized into three main questions: (1) If you accept my premise that there can be no mutually exclusive provisions in harmonized standards, then how can standards that contain different engineering standards (e.g., cage sizes) ever be harmonized? Can harmonization of the Council of Europe ever occur, based on the recently established cage size requirements? Compare those requirements with the provisions of the Guide. (2) Is it possible that you consider the general provisions of ETS 123 in the directive to be very similar to, even congruent with, the Guide, but that when you evaluate the specific wording, the two are different? If different engineering standards are permitted in so-called harmonized standards, may multinational companies (e.g., Dr. Kastello's company) allow the least stringent to meet the local requirements? Can they meet the requirements in Europe with the European requirements and in the United States with the US requirements? The preceding concern leads to the last question. (3) If the answer to question 2 is “yes, certainly they can,” what are the implications for the business operation of those organizations? Will they move their animal studies to locations with the least stringent standards, or will they, in fact, contract the work out to other countries with even less stringent standards?

DR. RICHMOND: I stated my position earlier as “yes, but”; however, when you have been married as long as I have, you know that “yes, but” is very similar to “no, unless.” To correct a misconception about the European situation, in fact, the Europeans will not be required to use particular cage sizes. In truth, the revised documentation will determine minimum sizes, but it will not dictate the size of the cage.

I think if you start from different positions, if there are economic as well as science and technical arguments to take into account, and if there is a lack of evidence on a number of the key provisions, it is necessary to work through a multi-staged process to achieve harmonization and to remain focused on the outputs. I think the first stage is to try to agree to the principles that should apply. I think everyone has agreed that good welfare and good science are worthy achievements. If we can agree to the principles, we might then be able to think about defining the processes that are needed to achieve those outputs—the mechanisms we might have to use, the judgments we might have to exercise, the evidence we might have to review. Those processes could be considered the second stage of harmonization.

If that level of harmonization achieves the desired outputs, we could stop. If after achieving that level of harmonization, no one has changed and everyone behaves as before (“because it has always been done this way”), I believe the regulators will decide to harmonize the practices. If the situation progresses to that stage, it will be the scientific community that has failed to adjust after the principles and the processes were harmonized. If it is possible to achieve the outputs through principles and processes, regulators have no reason to prescribe the practices.

DR. DELEEUW: I believe that discussions about harmonization of specific, quantitative matters such as cage sizes should be the last step of the process. In general terms, as Dr. Richmond said, we agree on the starting points, the principles. Where we might not always agree is on the steps between. We saw such examples in the descriptions of the Council of Europe proposals for rodents and rabbits, in the description of how much space is needed for certain types of enrichment. That example is the minimum freedom of movement that an animal should have. However, if we are able in the future to agree on basic principles, then the last step, harmonization of cage sizes, should not be a real problem. I believe it depends on the steps between, the translation of the principles to behavior.

DR. STOKES: I agree, and I think you have outlined a good process. However, it appears that we have moved in that direction without an inclusive process. Based on that premise, I would like to ask Drs. Richmond and de Leeuw the following question: Within Europe, with passage and adoption of these Council of Europe guidelines, is there the opportunity for exception, as Dr. Miller mentioned, to deviate and, in some circumstances, perhaps even use a smaller cage because there is a scientific rationale (e.g., to replicate data for a similar study that was conducted in the United States, with smaller cages)?

DR. RICHMOND: I think the answer is yes, that there will be the scope to do that. However, I am not certain of the actual specific arguments. The specific case you have described will not necessarily be accepted.

DR. DEMERS: When we are talking about cage size, we are talking more about standardization than about harmonization. I believe that we can agree on some specific concepts and some ideas. Some things depend on the culture (e.g., small hotel rooms in Japan vs. large hotel rooms in the United States—basically minor matters), and other things depend on economic purposes (e.g., exchange and trade); however, I do not know whether other things have an effect on science.

DR. STOKES: Dr. Gauthier recommended a process that would allow input from all of the involved stakeholders—the public, animal welfare groups, scientists, animal care specialists. Would any of the panelists like to propose how that process might be accomplished?

DR. GAUTHIER: I made that proposal because we already function this way in Canada. The process of guideline development involves first, of course, the scientific community, when we establish a subcommittee to develop a specific guideline; but also at the time of the second draft, we open it to public input. At that time, we also open it for other countries to have input. Most of the time, however, we ensure harmonization by including members of other countries in the subcommittee. This is the case, for example, with the subcommittee for fish and for other guidelines, when we are across borders with our southern neighbors and are taking action on the use of animals, where we have joint interface.

Also in our case, guidelines do not become standards per se until we have tested their rigor and solidity at the assessment level. On our peer-based team of assessors, we include representatives from the public who have been nominated by the Canadian Federation of Humane Societies, which entails additional input. We also have representatives on the institutional animal care committee, ranging from animal technicians to students. They all bring their input into the decision-making process and into these feedback learning loops that I mentioned. So it can be done. As we assess institutions and the basis of the guidelines, we find weaknesses that have been identified by peer-based assessment teams, which we later correct. We incorporate those items 2 years later as part of the review of the specific guideline.

This feedback mechanism is extremely useful. By the time we have completed the revision, we have already incorporated what the scientific community told us through the tour of our 190 participants in the program. There is no communication break per se; however, that continuity is actually part of the problem. The same thing is happening between countries, which is why I was calling for extended communication during guideline development with other countries, where it is relevant.

This mechanism should begin at some point. Then you see the value of it as you go through it. In our case, what it did for us was that the acceptance of the guidelines developed by everyone has reached a level now where government regulators are using the guidelines in areas where they have no specific jurisdiction to take action, or the public puts pressure on them to do it. We have extended the loop to government regulators, to the point where everyone is now included, including the regulators. So the process does work, with guidelines and standards.

DR. MILLER: Before we bring together a group of stakeholders, I believe we need to step back and ask why it is important. Why do we want to harmonize whatever we are harmonizing—guidelines or processes or whatever? What is the negative impact, and on whom is that impact negative, of not having harmonized guidance in this area? In other words, who cares? We could have a meeting like this, and we could all come together and agree on steps in harmonizing a process, but who would benefit? I think we have to answer those questions first, and know that there is a driving force for the effort. Perhaps Dr. Kastello would like to comment because he probably represents an affected group.

DR. KASTELLO: Certainly we would be affected by standards that would cause significant economic impact on the way we presently operate. Those of you who know me know that I am committed to animal welfare and to doing the right thing. As long as I have that responsibility in my organization, we will act accordingly. But I have to face the realistic problems of the economics. If the Council of Europe space revisions become part of the directive, they will significantly adversely affect my ability to house the numbers of animals I need to house for our present volume of studies in Europe. What does that mean to me? What are my choices? I can buy larger enclosures, I can build more facilities, or I can move the science somewhere else, where it is more cost-effective. Obviously, I will not be making those decisions, but we are talking about a great deal of money for one company. I speak only as a representative of a pharmaceutical company, but the entire industry will be faced with these kinds of decisions.

DR. GAUTHIER: I am reminded of what was said by one rapporteur this morning, which I believe is key. In assessing the value of any guideline or standard, rather than a single element like the measure of a cage, we should consider a program of environmental enrichment for the animals. The program should include, but not be limited to, cage size. This area is where we fail. We have a tendency to have a stovepipe approach, to limit what we think into one area. It is very scientific to try to exclude all other factors and retain only one factor, but the one we keep might be the one with the lesser significance. We should return to the concept that we are looking for and assessing a program, not simply the single element of a program.

DR. STOKES: At this point, I would ask for any comments or questions for the panel from the audience.

DR. ELLIOTT (Rosemary Elliott, Roswell Park Cancer Institute, Buffalo, NY): I would like to comment from the perspective of a mouse user. During this meeting, Dr. Smith presented data suggesting quite strongly that the current number of mice one is allowed to keep in a shoebox cage is very conservative. She estimated a number that was more than twice as large, which resulted in no major problems with the mice she was using, although admittedly the mice were a limited age range and a single strain. However, if these kinds of data are validated in the future, how will the various organizations respond? Can we put more than five mice in our standard cage? How will the European group respond? These are data. How are you going to work with these new data that will be coming out?

DR. GARNETT: I believe these data will clearly influence the existing processes, for example, what we hope will be a revision of the Guide in the next few years. The Guide revision committee should be looking at valid peer-reviewed science in formulating its collective judgment as to what is appropriate. So the more science, the better. I think many of these cage sizes we are dealing with now may not be supported by the hard science to which you are referring, although there was, obviously, consensus at some point in time. It is part of the existing process and should inform the decisions.

DR. MILLER: AAALAC has traditionally been a kind of “show me the data” organization. I believe there are some current, and I know some past, ILAR Council members here who might also address this issue. Dr. Golding, in fact, mentioned in one of the breakout sessions yesterday that AAALAC had identified what they considered a problem with three mothers. It was not a mandatory item, because we did not have the data. They kept three mothers and litters in one cage, and because it appeared that they were crowded, the suggestion was made to investigate, which they did. The resulting data indicated that the numbers were, in fact, detrimental; and they therefore discontinued that practice. Importantly, we are always seeking data. Dr. Smith's data—in her hands, in that strain, and under those conditions—were very compelling. I think it is relatively safe to say (although I never can speak for the Council on Accreditation) that if you were to reproduce for them Dr. Smith's data, they would find the data acceptable with respect to that density of animals.

DR. WITT (Clara Witt, Department of Defense, Global Emerging Infections, Silver Spring, MD): Based on my previous laboratory animal experience, I believe that Dr. Miller has brought up a very good point about the CIOMS. I have been a consulting veterinarian for the International Agency for Research on Cancer in Lyon, France—a United Nations (UN) organization animal facility that is supranational and not subject to the rules or regulations of any national jurisdiction. I too ask the following questions: What do we do with an animal colony? How do we apply for an NIH assurance letter for things like this? It was also necessary to work with the practicalities of everyday life—funding, good care of animals, what we do with our researchers, how we work with our researchers.

I agree with Dr. Miller that it may behoove us to look very carefully at the CIOMS principles, which served the UN animal facility very well. They provided the principles, the starting block for an international body of researchers, animal caretakers, and administrative staff, to work toward a continually improving animal care and use program. It was not perfect, and our animal care and use committees were “interesting,” to say the least, because various cultural, previous-research-background perspectives were involved. Yet the principles were solid, and we were able to work out many issues and eventually agree on some degree of standardization. For this reason, I do not use harmonization in the same sentence with standardization, because I think it blocks our ability to achieve harmonization toward the basic principle of good science and good animal welfare. I believe we all need to look at those principles very closely. It will be fantastic if at some point it is decided within the communities that those principles need to be updated and refined, although at this point they do provide a very good starting point for animal care and good research. [NOTE: See Appendix D.]

DR. MILLER: I am not certain whether the principles are online, but I propose including them on the ILAR web site. There are other multinationally agreed-upon guidance points, such as those of the International Conference on Harmonization, which involved Europe, Japan, and the United States, although those points are essentially restricted to pharmaceuticals. Dr. Richmond and I also worked together on an International Organization for Standardization (ISO) technical committee that has just completed (with Dr. Richmond's excellent leadership) a revised set of animal welfare standards: ISO standard 10 993, Part 2, Animal Welfare. There are approximately 80 members of that committee, and many countries have agreed to the very basic tenets. So if, in fact, it is a good thing to harmonize processes, we will at least have many resources and examples. The good laboratory practices of OECD, with which many people agree, also relate to animal welfare.

PARTICIPANT: As information, I would like to add that the CIOMS principles were developed under the auspices of CIOMS as part of the World Health Organization, and were chaired by a friend of all of us, Dr. Held. At that time, Dr. Held was chairing an NIH interagency committee, which redrafted the CIOMS principles into the US Government Principles, which you will find in the back of the Guide. These principles are the tenet of the Public Health Service policy.

DR. GARNETT: In addition, most of us in this room belong to professional societies, which also provide statements of principle or policy, or editorial review principles that paraphrase or restate almost verbatim the CIOMS and the US Government Principles. So they are already deeply entrenched in our minds.

DR. STOKES: I would like to add that those principles clearly incorporate Russell and Burch's principles of refinement, reduction, and replacement—the 3 Rs. These principles are clearly communicated.

DR. STEPHENS (Martin Stephens, Humane Society of the United States, Washington, DC): Based on the premise that animal research is a societal activity (i.e., currently more or less sanctioned by society, and regulated by governments that are more or less accountable to the people, at least in principle), then it follows that the stakeholders should be involved in whatever guidelines are written, whether it is within or across countries. Many of you have mentioned that animal welfare groups should be part of the process, and I believe their inclusion gives the process a certain legitimacy, because the result are guidelines that are carried out within the walls of research institutions, where it is difficult for the public to know what is happening. However, if animal welfare representatives are involved at the front end, there will be more assurance that the guidelines themselves are at a decent level.

We would like to see guidelines that have enough substance and specificity that there is some guarantee of a minimal standard. I do not personally like the minimal standard approach, but the public should be given some sort of guarantee that facilities cannot interpret those guidelines and basically do nothing. We know that some facilities are motivated to do the right thing and to do good, but we are particularly concerned about other facilities, which may have other priorities and allow standards to “fall through the cracks.” In addition, we are concerned not only about the drafting but also about the application of these guidelines. They should have enough substance so that facilities that basically flout them face some consequences, and so that the public has some guarantee that the system will ensure accountability.

DR. GAUTHIER: Dr. Stephens has stated a very important point regarding both the substance and the implementation of guidelines. I also mentioned this point in my presentation, when I talked about the Canadian mice that are more stressed than the American mice. I mentioned the recognition, during the June 2001 International Symposium on Regulatory Testing and Animal Welfare, that the ACC system is a very good system for implementing guidelines locally. Based on the work of the committee that Dr. Richmond chaired, most of the 22 countries now have reliable implementation systems in place that are being legislated by or based on a voluntary framework. We are not able to conceive and implement some harmonization measures, because we can trust that in all of those countries involved with harmonization, there is a system in which a third party can cross-check the implementation of the guidelines.

DR. BLOM (Harry Blom, Utrecht University, The Netherlands, and Vice President of International Liaisons, Federation of European Laboratory Animal Associations (FELASA)): I would like to mention an issue that, to my knowledge, has not been addressed in this meeting but that can have profound effects on the outcome of experimental results: how people have acquired their expertise and skills. As we all know, countries have different requirements for training and education, and when people move from one country to another, even though they have been trained, they often must complete additional training to practice in the new country. I would like to discuss this issue, and consider possible harmonization of education and training standards. I would like to hear from members of the panel on the issue.

DR. RICHMOND: Dr. Blom was too modest to declare a potential conflict of interest. Within Europe, FELASA has produced guidelines for key people within the animal care and use sector. Although the European institutions have not endorsed them, individual European countries have in most cases endorsed them. Within the United Kingdom, for example, we have our own training requirements, which are essentially theoretical. We do not allow practice on live animals (with one exception, which I will not detail at this time), and we require both basic skills and proper supervision and training in bona fide testing and research programs. We are uneasy with people having boxes of practice animals for practice in facilities.

We recognize relevant training from any other European countries, provided there is documentary evidence, and we have one additional requirement—completion of module 1, which is essentially an overview of UK domestic legislation. We do not expect a visitor from overseas to be equipped with that information in advance. We have recently accredited an Internet distance-learning course that will allow people to undertake the training before they arrive.

DR. MILLER: Dr. Blom's suggestion is excellent. As you know, AAALAC already has included the FELASA guidelines in our reference resources. In the United States, the IACUC must ascertain and confirm that everyone who is doing something with animals is appropriately trained and qualified. It is a very broad performance standard.

DR. GAUTHIER: In Canada, we have been promoting the training programs given by institutions. In 1999, we developed and published a guideline on the training of animal users, which became mandatory for assessment purposes in January 2003. We are now verifying, through our regular 3-year cycle of assessments of all institutions, that institutions do have the training structure in place. In addition, we provided assistance by publishing on our web site 12 modules on training for animal users— mainly researchers, graduate students, and research staff. The modules are a very good continuing education tool, which is available in both French and English. It is a mandatory requirement, but hands-on training, of course, is the responsibility of the institutions as per our guidelines.

DR. NELSON (Randall J. Nelson, Department of Anatomy and Neurobiology, IACUC Chair, University of Tennessee, Memphis): It is my responsibility to share these discussions with fellow IACUC members, and I will say that we agree on many issues. However, I submit that we still may need to deal with another aspect of implementation philosophy. I would describe the two sides of the issue as centralized versus decentralized views. To use the analogy of our own conflict about 140 years ago in the United States, these views might be analogous to states' rights versus federalism. I have heard that some implementations are very centralized, having a set of rules, guidelines, or prescriptions that must be accomplished by all involved, and they are being implemented at a national or supranational level. I have also heard of an implementation structure that seems to be diffuse or distributed to individual IACUCs and/or institutions, allowing those bodies to set certain internal standards, provided they do not deviate from the Guide.

In both cases, I can see benefits, for example, consistency in the centralized system versus markedly different implementations, presumably, at different institutions. I would like the panel to discuss the benefits and the drawbacks of these two systems. I believe it will be necessary to deal with those differences in philosophy before we can decide whether we can harmonize or standardize.

DR. MILLER: Dr. Nelson makes an excellent point and has identified the advantages to both of those types of systems. As I mentioned earlier, I view AAALAC as performing de facto harmonization, or implementation. I authored an article in Lab Animal in the late 1990s titled “Harmonization: The Proof Is in the Practice.” In retrospect, the title should have been “The Proof Is in the Performance.” You can have all of the desired harmonized standards and all of the different desired standards, but these desires are meaningless unless you monitor implementation. You have no system unless you have a way to verify that an institution has fulfilled its national requirements.

I take pride in AAALAC because in more and more countries, we assess performance based on a host of standards that begin with the respective national requirements. The basic principles of the Guide must always be met. You can be assured that our evaluation of performance under those circumstances is consistent to a great extent, that what takes place in an institution we accredit in India is very similar to what takes place in an institution we accredit in Indiana. So I believe it is possible to establish consistency. There are advantages and disadvantages to both, but if you want to harmonize process, you need some sort of national system. You do not need regulations, as Canada has shown us.

For those countries that do not already have a system in place, the Canadian system is absolutely the best, bar none. You start with some sort of national system. Then you establish a responsible authority, which includes a process for reviewing the ethics and the procedures of each proposal. The authorities can be in the same body or they can be separate bodies; they can be national, regional, local, or whatever—as long as they exist. Finally, you need a monitoring mechanism to confirm that requirements are being met. You can have all of the rest, but it will be a hollow exercise unless you have that last mechanism—whether, as in the United States, it is the Department of Agriculture coming in and doing unannounced inspections, or a voluntary program like AAALAC.

DR. GAUTHIER: The Canadian Council on Animal Care operates a decentralized quality control peer-based system through the institutional animal care and use committee. This system assesses the functioning of all of those committees as well as the facilities and so forth. The same people who are part of their own institutional animal care committee, at a different time, become assessors on one of our teams. They contribute by generating and circulating the knowledge from local to national, and from national to local levels. It is, again, decentralized quality control, or centralized quality assurance.

DR. RICHMOND: I believe that some things are best done, perhaps should only be done, at the top—things like policy and principles probably must be top-down, rather than bottom-up. However, processes and practices must be bottom-up. I cannot sit in an office 380 miles away from a research laboratory and know what is best for them to do in the context of what they are doing that day.

With regard to verification, I would like to repeat three key points: (1) If you are spending government money, and if you are at the user end, you must be absolutely clear about what you are actually being asked to do or what you have actually contracted to do. (2) You must have an intelligent regulator who has the degree of technical expertise to argue the technical points with you, rather than simply restate the regulation. (3) You need to add to your oversight system a mechanism that draws attention to things that are now going out of date and need to be replaced, and something that highlights the area where people have excelled and produced something that may be better than the common current standard.

PARTICIPANT: I would like the panel to suggest ways to close the knowledge gap, to generate the necessary data that we have discussed during this workshop. I would suggest that we consider and possibly adapt for our needs an existing model operated by the organization called International Life Sciences Institute (ILSI), which addresses various public health questions. When they have a question, they convene a meeting of the stakeholders who are interested in that question who then form a consortium. They look at the work that needs to be done, they divide it among themselves, and they then perform different experiments. The protocols are reviewed by a central steering committee, and the results are shared with the organization and eventually published. That system is one possible model for the laboratory animal community to use in addressing questions on animal welfare. We have many different strains and species, and more unknowns than knowns. One possible way of attacking the problem, rather than simply scheduling a meeting 5 years from now and admitting that there is still a dearth of knowledge, is for one of the nonprofit international organizations to step forward and entreat people to participate in a consortium, divide up the work, and start doing it. Stakeholders not only could do some work on their own, but a body could apply for public funding by forming such a consortium. Universities and other institutions that are not profit-making institutions could then participate in the process as well.

DR. KASTELLO: Some organizations provide funding in this area, for example, we heard from Ms. Cathy Liss that the Animal Welfare Institute is providing some funding to work in this area. Since 1996, the American College of Laboratory Animal Medicine Foundation has provided small research grants to work at filling these gaps. It is not a huge amount of money, but it is at least an effort to begin to understand more about the animals we are using. Clearly, that process needs to be improved, and it needs to be advertised more widely so that people know there are some funds available, and to encourage people to submit research proposals.

DR. MILLER: What is the source of the funding in the ILSI program?

PARTICIPANT: The stakeholders provide the funds. The companies themselves pay for the studies in which they are most interested. The studies are prioritized, the work is divided, and the companies participate in doing that work.

DR. MILLER: It sounds like an excellent model. We are always left with the question, where will the money come from?

DR. KASTELLO: The European College of Laboratory Animal Medicine has also started a foundation. FELASA has helped fund it. The intent is to fund similar research grants, to enhance the knowledge about the animals we use.

DR. GARNETT: I would like to share with you briefly my meeting with NIH Director Dr. Zerhouni, when he first went to NIH. In response to his request for my (and other office directors') highest priority, I described the need for funding to answer scientific questions to inform public policy in the animal welfare area. He was very interested and asked for examples. Among the examples I cited of questions for which we have insufficient data—questions for which we need scientific answers—was whether decapitation is humane. I also cited issues related to the appropriate uses of CO2. I believe he took my responses very seriously, but he also needs to hear from people other than his own staff. I believe animal welfare is a worthy thing to communicate, and you may agree that this issue should be a priority for NIH to address.

PARTICIPANT: I think all of the funding sources mentioned are very helpful, but the problem I see is a lack of coordination. If you had a consortium that listed the work to be done, prioritized it, and then used all available funding sources, you could get much of the work done in 3 or 4 years.

DR. STOKES: In closing, I would like to thank all of the panelists for their participation and for preparing remarks on their initial positions regarding harmonization. Clearly, our discussions reflect that we share common ground on one aspect of harmonization—the international principles on the humane care and use of laboratory animals. It appears that we have some work to do regarding the harmonization of processes, standards, and practices, but I think that opportunity is in the future.

DR. KLEIN (Hilton J. Klein, Merck Research Laboratories, West Point, PA): I would like to echo Dr. Stokes in thanking all of the participants. The dynamic exchange that I have perceived over the last 3 days has been incredible. In case I am an example of anyone else in this room, I will mention that I came here with some trepidation and anxiety, and I have heard many discussions and comments about people being pleasantly surprised. I think one of the results of the workshop was to open key channels of communication, and we have learned that some of the perceived gaps are not as real as we had thought.

In addition to thanking the participants, I sincerely thank all of the speakers, and particularly our plenary speaker Dr. Crabbe, our banquet speaker Dr. Holekamp, every other speaker on the program, and especially members of our panel, who did a marvelous job. I think the sharing of knowledge was facilitated by the fact that we planned the workshop to be interactive and participatory, and I believe the breakout sessions enabled us to achieve that goal. In summary, from the perspective of ILAR, this meeting has been one of discovery. All of it has been about science. Science is defined simply as knowledge. The English word is derived from Latin, which denotes the systematic observation of natural phenomena for the purpose of discovering laws governing those phenomena. This workshop certainly was a phenomenon.

Some issues that arose were surprises, for example, the fact that we do not know our own regulations and our own country's standards as well as we should. We need to improve in that area, and I think this workshop is a step in that direction. We heard that message repeatedly. I would also like to emphasize that we all need to communicate these issues to the scientific community. We need them on our side to address the issues. I think we all have the combined common goal of improving welfare and performing good science, but we need the scientists in the room. We have had some discussions about how to accomplish that goal. The first thing to do when you return to your laboratories is to share this information with your colleagues. This issue is very important both today and for the future, because if we do not resolve the problems ourselves, the public will expect others to do it for us. I do not think any of us want to be in that position, because the privilege of using animals in research, worldwide, is at stake. To maintain that privilege, we must address the issues around it. Our understanding of the guidelines, what drives them, what creates them, the science behind them, and where we get the funding—all of the issues we have discussed—will preserve and protect that privilege and the stewardship that is part of using animals in research. On behalf of the ILAR Council International Committee, I again thank every one of you for making this a wonderful experience.

DR. DEMERS: I have been asked by my colleagues at the international level to express our many thanks to ILAR and to the National Academy of Sciences, to Dr. Hilton Klein, Chair of the Program Committee, and to all members of the Program Committee. We also thank Ms. Kathleen Beil and all of the support staff who have assisted us in the planning of this meeting. We thank those who were involved in the organization of the meeting, and especially ILAR Director Dr. Joanne Zurlo. Thank you for providing us with the opportunity to explain and describe how we see things. We all are seeking the same high-quality result regarding the welfare of animals, and we all are scientists. Finally, but very importantly, we thank the sponsor and cosponsor of this workshop, without whose support its success would not have been possible.

I hope that this initiative will be repeated. I urge all of you to attend the next FELASA meeting, to be held next year in Nantes, France, where the theme of the meeting will be internationalization and harmonization in laboratory animal care and use issues. This meeting will provide an opportunity to maintain the ongoing dialogue, because even if we agree or disagree on harmonization, I believe that most of us agree that it is important to keep communication active and proactive.