Summary

Dogs were used in biomedical research as early as the 1st century CE and continue to be used as laboratory animals to the present day, due to their anatomical and physiological similarities with humans, and likely because of their ubiquity and comfort around humans. Today, many public and private institutions in the United States use animals for biomedical research purposes, but over the years the use of dogs has declined. Although the use of animals in biomedical research is regulated by federal laws and is subject to institutional oversight to ensure the humane use and care of laboratory animals, this practice remains contentious and a focus of intense public scrutiny, particularly in the case of dogs.

STUDY ORIGIN

At the request of the U.S. Department of Veterans Affairs (VA), the National Academies of Sciences, Engineering, and Medicine (the National Academies) convened an ad hoc committee to conduct an independent assessment of whether dogs are or will continue to be necessary for biomedical research directly related to the VA’s mission. The committee addressed the care and use of laboratory-housed dogs but not of privately-owned companion dogs enrolled in clinical trials. The entire charge to the committee is shown in Box S-1. Note that non-human primates were excluded from the committee’s consideration according to the Statement of Task.

To address the study’s Statement of Task (see Chapter 1), the National Academies appointed a 16-member committee with expertise in laboratory animal veterinary medicine, biomedical science, medicine, translational research, animal welfare science, animal behavior, animal welfare regulation, alternatives development, bioethics, and animal law.

As one component of its effort to understand the procedures governing animal care and use at the VA and the scientific justification for undertaking research in laboratory dogs, the committee reviewed 14 animal component of research protocol forms associated with current or recent past studies proposing the use of laboratory dogs, as well as additional background materials provided by the VA. The committee supplemented information from expert panels with a comprehensive

review of the current scientific literature regarding clinical questions related to the VA’s mission that have traditionally used dogs, taking into account both animal and non-animal alternatives to laboratory dog use (see the literature search parameters in Appendix A).

LEGAL, SOCIAL, AND ETHICAL CONSIDERATIONS

Conclusion 2-1: Based on the documentation provided by the U.S. Department of Veterans Affairs (VA) and other organizations (e.g., confirmation of AAALAC International accreditation) as well as site visits, the committee concludes that the VA’s biomedical research programs involving laboratory dogs appear to adhere to all relevant policies surrounding animal research.

There is evidence that dogs were the first animal to be domesticated, deriving from the grey wolf at least 14,200 and perhaps as long as 36,000 years ago.¹ Unlike other domesticated species such as pigs and cows, which were primarily selected for traits related to food production, the ancestors of dogs were initially selected for their behaviors.² The ancient bonding relationship

___________________

¹ Ovodov, N. K., S. J. Crockford, Y. V. Kuzmin, T. F. G. Higham, G. W. L. Hodgins, and J. van der Plicht. 2011. 33,000 year-old “incipient” dog from the Altai Mountains of Siberia. PLOS ONE 6(7):e22821.

² Serpell, J. A., and D. L. Duffy. 2014. Dog breeds and their behavior. In A. Horowitz (ed.), Domestic dog cognition and behavior. Berlin, Heidelberg: Springer-Verlag. Pp. 31–57.

between Homo sapiens and Canis lupus familiaris, which is manifested in the dog’s DNA, is arguably unique. One feature of this relationship meriting reflection is that the docility of dogs, along with their willingness to obey commands and trust humans, makes them easier to work with as laboratory subjects than other large animals. Today, dogs are currently seen in the United States as service, therapy, working, and companion animals—which is likely one driver of public resistance to the use of dogs as laboratory animals.

The preference for dogs can also be viewed as prejudicial. That is, insofar as honoring such a preference would substitute other species for dogs, it discounts the welfare and interests of other species merely because many individuals prefer dogs. Most accounts of animal ethics reject “speciesism” (i.e., the notion that it is acceptable to discriminate against some animals based on their species membership), or at least they place strong constraints on it. The moral relevance of the human–dog unique relationship is highly debatable. According to prevailing accounts of animal ethics in the scholarly literature, there is no basis for discounting the interests and welfare of animals like sheep or pigs relative to dogs.

A condition stipulated by some ethicists is the notion that no experiment should entail unnecessary harm, where necessity is dictated by the purpose of the experiment. On this view, using a species that would experience greater welfare loss for a given research objective would count as unnecessary harm. Rejecting speciesism does not logically entail that all animals should be treated exactly the same. Different animals require different things to flourish. Some animals need certain kinds of bedding, others need certain forms of mental stimulation, and others need space to roam. A rejection of speciesism entails that equal consideration be given to the equivalent interests of different animals and members of different species.

Conclusion 2-2: Many people have a unique relationship with dogs. This relationship stems from tens of thousands of years of joint history and the cultural value of the role of dogs as service, military, law enforcement, and working partners as well as companion animals. This cultural preference for dogs is not universal, nor does it necessarily constitute a reliable guide to ethical action. The majority of the committee concludes that it is valid to consider the societal preference for dogs only in situations where expected burden for substitute species is anticipated to be equivalent to that projected for the laboratory dog.

DETERMINING THE NECESSITY OF LABORATORY DOGS IN BIOMEDICAL RESEARCH FUNDED BY OR CONDUCTED AT THE VA

The committee considered dog use in 10 biomedical research fields related to the VA’s mission (see Chapter 3). Cardiovascular disease, spinal cord injury, and imaging are areas of current VA biomedical research using laboratory dogs;³ diabetes, narcolepsy, osteoarthritis and chronic pain, and experimental pharmacology and toxicology are areas of recent past VA research; and the committee identified three areas considered to be potential candidates for future VA research using (laboratory or companion) dogs—cancer, infectious disease, and Alzheimer’s disease. It would not have been feasible for this committee to cover all possible research areas of current or future interest to the VA, and the absence of a particular field from this report should not be taken as a determination regarding the necessity of dog use in that field.

___________________

³ A list of currently active studies in the VA biomedical portfolio that use dogs is available online at https://www.research.va.gov/programs/animal_research/canine_research/current_research.cfm (accessed December 10, 2019).

Recommendations 1, 2, and 3, if adopted and enforced, would become part of the culture and process of scientific and ethical review at the VA. Recommendations 1 and 2 would create the expectation that principal investigators consider, early in the study proposal process, all possible alternatives (non-animal or animal) and the relative harm the proposed study would bring to the candidate subjects. Scientific review committees and institutional animal care and use committees (IACUCs) would be conducting simultaneous reviews of the analysis of harm and benefit, such that all three parties—principal investigator, scientific review committee, and IACUC—would develop an agreed-upon understanding of “scientific necessity,” reconcile any differences of perspective related to the proposed study, and generally pool accountability for decisions related to the use of laboratory dogs.

After reaching agreement on Conclusion 3-1 and Recommendation 1, the committee found itself at an impasse. Ten committee members,⁵ a majority, believed that according to the Statement of Task their job was not done and that a second recommendation linked to Recommendation 1 was warranted, while five committee members⁶ were equally convinced that this second recommendation would not be in keeping with the Statement of Task and thus should not be included in the report. The differing opinions of the two groups turn on the meaning of three specific sentences in the Statement of Task—and, in particular, on the meaning of one word—“necessary”—that appears multiple times in those sentences.

To understand the disagreement, some background is useful. Throughout the study process, the committee debated at great length how scientific, legal, ethical, and social considerations factor into the determination of the laboratory dog’s necessity in VA biomedical research. In March 2018,

___________________

⁴ Text was modified after the release of the prepublication report to the sponsor to clarify that some of the criteria in Recommendation 1 are not new to the VA. The committee intends for the criteria, old and new, to be applied as a complete set.

⁵ W. Ron DeHaven (Vice Chair), Joan Hendricks, Jonathan Kimmelman, Lewis Kinter, Nancy Marks, Christian Newcomer, William Potter, David Powell, Margaret Riley, and Rodney White.

⁶ Rhonda Cornum (Chair), Donna Arnett, Warren Casey, Chris Green, and Sarah Lathrop.

prior to the VA’s request that the National Academies undertake this study, the federal government enacted new restrictions on the VA’s use of laboratory dogs, mandating that no federal funds, “may be used to conduct research using canines unless: the scientific objectives of the study can only be met by research with canines.”⁷ Section 254 of the Consolidated Appropriations Act of 2018 further required the Secretary of the VA to “directly approve” any such studies, and to submit to the U.S. Congress within 180 days, “a detailed report outlining under what circumstances canine research may be needed if there are no other alternatives.”⁸

In December 2019, as this committee neared the end of its deliberative process, the Further Consolidated Appropriations Act of 2020 was enacted into law. This legislation reiterated the language from the 2018 Act and expanded it to include cats and non-human primates.⁹ The 2020 Act also added a new requirement that such scientific objectives must be “directly related to an illness or injury that is combat-related.”¹⁰ Furthermore, the 2020 legislation now requires the Secretary of the VA to submit a report to the U.S. Congress for any such approved research, “not later than 30 days before the commencement of such research.”¹¹ That report must describe the nature of the research and include “the justification for the determination of the Secretary that the scientific objectives of such research could only be met using canines, felines, or non-human primates.”¹²

In considering the Statement of Task in the context of this legislation, the members of the majority and minority disagreed about the scientific and ethical implications of this legislation and about its relevance to the committee’s recommendations to the VA. Chapter 3 contains a description of the areas of disagreement. To sum up, the disagreement between the majority and minority over Recommendation 2 is essentially a disagreement about whether that recommendation comports with the Statement of Task. The majority, taking a broad view of the meaning of “necessary,” believes it does. The minority, holding a more restricted view of the meaning of “necessary,” believes it does not. The practical effect of that definitional disagreement is that the majority believes that the interests of other laboratory animals than the dog must be taken into consideration when determining the necessity of research on laboratory dogs, while the minority believes that the question the committee was asked dealt not with other research animals but only with laboratory dogs.

Recommendation 2: Adopt an expanded set of criteria for determining when to use laboratory dogs in the U.S. Department of Veterans Affairs’ (VA’s) biomedical research when the dog is not scientifically necessary.^13,14

___________________

⁷ Consolidated Appropriations Act of 2018, Sec. 254, p. 825 (U.S. Congress, 2018).

⁸ Id.

⁹ Further Consolidated Appropriations Act of 2020, Sec. 249(a)(b), pp. 665–666 (U.S. Congress, 2019).

¹⁰ Id., p. 666.

¹¹ Id.

¹² Id.

¹³ Five committee members (Rhonda Cornum [Chair], Donna Arnett, Warren Casey, Chris Green, and Sarah Lathrop) dissent to Recommendation 2. The dissenters acknowledged the English dictionary definition of “necessary” (“required to be achieved, or essential”) as outlined in the Statement of Task to recommend to the VA when the laboratory dog was necessary in biomedical research (i.e., the dog is the only model that will yield scientific results directly related to veterans’ health). This is exactly what federal law currently directs. As the Statement of Task did not request an evaluation of other animal models, the dissenters conclude the majority’s Recommendation 2 strays beyond the Statement of Task. Additionally, the five committee members argue that a broader ethical framework that is responsive to the public’s perception of animal research be considered, especially given that research conducted by the VA is publicly funded.

¹⁴ Text was modified after the release of the prepublication report to the sponsor to clarify that some of the criteria in Recommendation 2 are not new to the VA. The committee intends for the criteria, old and new, to be applied as a complete set.

• Protocol Development and Review Processes. Specifically, the VA should:
  • Emphasize the replacement of laboratory dogs and the refinement of procedures and techniques over a reduction in animal numbers in order to reduce the burden on individual dogs, even if more animals (including alternative species) will be used;
  • Improve literature searches for alternatives to laboratory dogs. The VA should fund an independent party to conduct literature searches designed to yield objective, independent analyses of the need to perform proposed research in laboratory dogs versus alternative animal models, NAMs, humans, or human tissues; and
  • Engage with board-certified and other experts in canine medicine and research to review research goals and ensure optimal study design, including estimates of the sample size needed to ensure adequate statistical power. Consider spontaneous clinical conditions of relevance and the possibility of clinical trials in companion dogs to complement or replace laboratory dog studies.
• Track Impact of Research. Specifically, the VA should:
  • Establish a mechanism for tracking the impact and translation of research using dogs. Such a retrospective reporting mechanism should use objective and state-of-the-art methods (e.g., bibliometrics or citation in regulatory documents and patents) to track the relationship between dog experiments and translated interventions for veterans. Such performance assessment should be required to establish and, if need be, correct risk–benefit and welfare assessments used in the authorization of research.
  • Take steps to encourage the prospective registration of all studies involving laboratory dogs.

ALTERNATIVES TO THE USE OF LABORATORY DOGS

Shifting research to an alternative model (animal or non-animal, including humans) will require successfully addressing those factors that favor the continued use of laboratory dogs. The VA has an opportunity to become a premier biomedical research entity engaging formally with veterinary expertise, both to enhance the experience of laboratory dogs and to conduct clinical trials in companion dogs, using companion dog studies to replace laboratory dog research wherever possible. Accomplishing this goal will require the VA to do the following: (1) engage with experts in canine medicine and research to optimize both clinical methods and research goals, (2) collaborate with researchers conducting clinical trials in companion dogs to identify or develop trials to benefit veterans, and (3) participate in efforts to develop a registry connecting human research needs with companion dog clinical trials.

could be identified and if there is an infrastructure in place to facilitate the conduct of studies that use companion dogs.

Conclusion 4-2: A significant barrier to conducting clinical studies in companion dogs is a lack of administrative infrastructure to connect U.S. Department of Veterans Affairs (VA) investigators to the veterinary researchers who conduct such trials. The regulatory infrastructure to address ethical and legal issues for clinical trials in dogs is already established, but the mechanism for using these studies to supplement, complement, or accelerate collaborations with investigators who are interested in conducting human clinical trials does not exist. With validation of the utility and relevance of the naturally occurring canine disease or disorder for the study of the human equivalent disease or disorder, a network for developing a companion animal clinical trials registry could be created. The VA could move forward with supporting new collaborations to establish the relevance of dog studies to humans, both for conditions of likely future interest to the VA (e.g., posttraumatic stress disorder, natural infectious disease, Alzheimer’s disease, and obesity) and for areas currently under study.

Conclusion 4-3: With respect to other animal models, rats and mice are the predominant species used for biomedical research in the fields of cardiovascular disease, spinal cord injury, cancer, and diabetes. For studies that cannot be performed in rodents (due to constraints of size, anatomy, or physiology), the pig has become the large animal translational model of choice. While pigs are not tractable for all areas, their potential uses are likely to expand in the near future as genetically modified strains become more widely available.

Conclusion 4-4: While the scientific and institutional animal care and use committee review processes at the U.S. Department of Veterans Affairs adhere to all relevant policies established by the U.S. government (as described in Chapters 2 and 5), compliance with these standards on its own may not be sufficient to ensure adequate identification and consideration of new approach methodologies (NAMs). Even in the case of protocols that still require the use of laboratory animals, researchers need to be encouraged to evaluate and incorporate NAMs where feasible.

Recommendation 4: Develop a strategic roadmap to create, track, and sustain internal efforts to incorporate new approach methodologies (NAMs) in U.S. Department of Veterans Affairs (VA) biomedical research.

The VA should establish a strategic roadmap and accompanying framework to promote the development and incorporation of NAMs to replace, reduce, or refine the use of dogs and all other laboratory animals in VA research. This framework should prioritize:

• Modifying the protocol review processes (see Recommendation 3) to require and support robust consideration of NAMs, human clinical trials, companion dogs, and alternative animal models. The potential of these alternatives to contribute to the overall goals of the research, not just to replace laboratory dogs, should be considered.
• Incentivizing the use of NAMs. Examples of ways to do this include:
  • Developing and funding new VA grant opportunities to promote the development of NAMs that meet the unique needs of VA researchers, including the use of human tissues and organs, in vitro, in silico, and computational approaches.

• Funding for researchers and institutional animal care and use committees to undertake training in state-of-the-art, human-based methods to increase awareness and help establish confidence in these new approaches. Hands-on training and similar knowledge transfer opportunities will be particularly important and should be prioritized.
• Implementing compulsory funding to promote the evaluation and optimization of NAMs that address research objectives identified in studies that currently require the use of laboratory dogs (i.e., parallel funding requirements).

Recommendation 5: Establish long-term external collaborations to optimize the use of companion dogs and humans in biomedical research.

The U.S. Department of Veterans Affairs should prioritize the development and continuation of external multi-disciplinary collaborations to develop, validate, and apply alternatives to the laboratory dog in biomedical research. This effort should result in the following:

• Increased collaborations with external scientists and use of public–private partnerships to promote cross-sector communication and cooperation.
• The fostering of collaborations with researchers conducting clinical trials in companion dogs to identify or develop trials to benefit veterans and dogs.
• The encouragement of the use of human organs and tissues from human organ banks whenever possible.

CARE AND WELFARE OF LABORATORY DOGS USED IN RESEARCH

Over the past 50 years, an impressive array of work has been done to understand the welfare of animals in a range of settings, from farms and laboratories to zoos, shelters, and the wild. To survey that work, even within the laboratory setting, is beyond the scope of this report, but four key conceptual advances from this period inform practice regarding the welfare of animals in human care today. These advances reflect developments in the field of animal welfare related to the following and are discussed in more detail in Chapter 5: animal sentience, the emergence of three distinct approaches to the nature and assessment of animal welfare, a consideration of positive and negative welfare states, and the recognition that welfare assessments incorporate both resource-based and animal-based considerations.

Conclusion 5-1: Animal welfare is multi-dimensional, reflecting the health, comfort, behavior, and emotions of animals in human care. There is a need for the U.S. Department of Veterans Affairs to combine aspects of each of the three major approaches to animal welfare, explicitly consider positive and negative welfare states, and measure animal-based welfare indicators in order to enhance the positive welfare of laboratory dogs.

Conclusion 5-2: While the Guide for the Care and Use of Laboratory Animals and the Animal Welfare Act provide a foundation for the assessment of well-being, the U.S. Department of Veterans Affairs has an opportunity to incorporate current developments in animal welfare science into its animal care and use program.

Conclusion 5-3: In the pursuit of enhancements to the care and use of laboratory dogs in U.S. Department of Veterans Affairs (VA) research, international regulations and guidelines may

be a useful resource for alternate approaches and recommendations. International guidelines and recommendations offer the VA programmatic leaders and institutional animal care and use committees expanded perspective on the contemporary literature and trends on laboratory dog care.

Some members of the committee conducted visits to two of the VA programs engaged in dog research in order to meet with personnel involved in oversight, care, and research on dogs and to observe the conditions and practices associated with this use. Materials obtained in advance of the National Academies site visit to the Richmond VA Medical Center indicated that practices relating to dog husbandry and care were sound and conformed with the applicable requirements. These materials included two letters of accreditation from AAALAC International, which strongly supported the VA’s assertion of a high-quality program for the care and use of dogs. External review by the National Institutes of Health’s Office of Laboratory Animal Welfare and upper management within the VA system also found provisions for the care of dogs at the Richmond VA Medical Center to be in compliance. Both the Richmond and St. Louis support facilities, such as surgeries and treatment rooms, were state of the art and well maintained.

• • When it is compatible with study goals and safe for the dogs and personnel, create an outdoor space for laboratory dogs to visit on a regularly scheduled basis. This would provide additional opportunity for exercise as well as olfactory, sensory, and visual stimulation; a variety of experiences; and time with humans.
  • Increase the amount of enrichment available to dogs, and continue to evaluate and incorporate new options for environmental enrichment, including olfactory enrichment, on a regularly scheduled basis.
• Address current experimental impediments to dog–dog interactions.
  • Given concerns about possible wounding or damage during the social housing of dogs wearing internal or external (implanted) devices, fund a pilot study at the end of an existing protocol to examine the risk of these outcomes.
  • Consider an alternative placement of implanted devices to decrease the likelihood of complications from socialization with other dogs.
  • Encourage the development and use of miniaturized devices that are less cumbersome for the animals and less likely to be damaged, especially if they can be implanted subcutaneously.
• Conduct enhanced assessments of laboratory dog welfare.
  • To move beyond simple observations of dogs’ health, VA staff involved in the care and welfare of laboratory dogs should collaborate on continuous education and continuous improvement of measures that advance laboratory dog welfare.
  • VA veterinary and animal care staff, facilities personnel, members of the institutional animal care and use committee, and principal investigators should conduct formal, written assessments of animal welfare that reflect the state of the art in animal welfare assessment methods.