CARINE ELKHORAIBI, AMY ROBINSON-JUNKER, GINA ALVINO, AND LARRY CARBONE

Institutional Animal Care and Use Program, University of California, San Francisco, California, USA; Laboratory Animal Resource Center, University of California, San Francisco, California, USA; Institutional Animal Care and Use Program, University of California, San Francisco, California, USA; Institutional Animal Care and Use Program, University of California, San Francisco, California, USA

The Health Research Extension Act of 1985 is the AWA’s companion law. It requires National Institutes of Health oversight of all vertebrate animals in research, based on compliance with the Guide for the Care and Use of Laboratory Animals (the Guide), though only if the work is funded by federal research grants (Institute for Laboratory Animal Research, 2011). Animals these laws do not cover, such as mice in private industry or in undergraduate classrooms, may be covered by state or local laws or voluntary accreditation. Or, their care may have no external oversight at all.

Russell and Burch (1959) developed a widely used framework for improving laboratory animal welfare. They described a system for categorizing the sources of pain, distress, and suffering. Suffering can be the result of how scientists house and care for research animals, a result of how they use them in experiments, or both. Animal suffering in laboratories can be either direct or contingent. Animals experience direct suffering when their suffering is the subject of study. For instance, a researcher investigating the neurobiology of pain may need to first induce pain; prevention or alleviation of that pain would invalidate the study. Contingent suffering is a side effect of the animal’s environment or experimental use, such as pain after a surgical procedure.

Along with a framework for identifying causes of suffering, Russell and Burch also proposed their framework for reducing suffering: the ‘Three Rs’ of replacement, reduction, and refinement alternatives. Replacement refers to replacing sentient animals with nonsentient cells, computer simulations, even human clinical data. Reduction comprises the various ways an experiment might use fewer animals. Refinement is the suite of activities that can improve humane treatment when animals must be used. Most of our coverage in this chapter pertains to various refinements of animal housing and animal use.

Suffering is not purely pain or exclusively physical. Regulations have increasingly and incrementally focused explicit attention on the mental health of laboratory animals, but this was not always the case. The US Drug Administration (USDA) once feared that anxiety was a ‘psychiatric term that is only applicable to humans’, while distress is ‘more descriptive of the physical visible state of the animal’ (Carbone, 2004, p. 209). Then Congress amended the AWA in 1985 with a provision for environments that enhance the psychological well-being of zoo and laboratory animals, though only for nonhuman primates (NHPs).

The Guide has also expanded its focus on the mental health and well-being of animals in laboratories through its eight editions. It now sets some general standards for behavioral management and for enrichment, including a default expectation that social animals should be housed with compatible conspecifics (Institute for Laboratory Animal Research, 2011). For the most part, welfare standards in the Guide and the AWA are quite general and leave to individual research institutions, working with their institutional animal care and use committee (IACUC) and their veterinarians, the development of their own specific policies and procedures.

19.3 Animal Welfare and Animal Research Data: How One Influences the Other

At first this seems paradoxical: most biomedical animal research requires healthy, normal, comfortable, unstressed animals. The researcher wants her animal groups to differ solely by whether they are receiving the placebo or the test cancer drug. She wants them to be eating normally, not stressed by rough handling or painful injections. Their cage should be comfortable, not too cold or hot or stuffy or brightly lit. If they live in groups, the groups should be compatible, with no fear of bullying, no fighting, and no painful infected fight wounds. If the drug makes the animals too sick or in too much pain to eat, it introduces unwanted variability in nutritional status. If cancer rates then differ, is this credited to the drug or to nutritional factors? Have animal lives been wasted on an inconclusive experiment?

Russell and Burch used the term psychosomiasis to describe the very real interaction between physical and mental health, a mind–body interaction that informs this chapter (though modern writers avoid this term and its modern connotations; see also Chapter 4, this volume). Significant mental and emotional stressors can lead not just to distress and poor mental well-being in animals, but also to physical consequences. Severe stress may lead to illness, which may in turn add to the animals’ mental suffering. Stress may also have less dramatic effects on the animal’s body, altering immune function, for example, but not to the degree that the animal actually feels ill.

Knowledgeable animal researchers understand that changes in an animal’s emotional or affective state can affect how the animal’s body functions, even if they do not know the specific physical effects of specific changes in animals’ mental state. Scientists evaluate any suggestions laboratory animal professionals or IACUCs make to improve the animals’ welfare through the lens of their concern over outcomes in physical function and the quality of their data.

While some welfare improvements raise concerns about the quality and variability of the effects on data, animal welfare improvements may instead lead to better science. For example, the less stressed the animal is, the more normal many aspects of the animal’s biology and the more reliable studies of that animal may be. Unfortunately, there are times when welfare improvements can work against the goals of the experiment. For example, mice on a lung cancer study might feel more comfortable with supplemental oxygen therapy, but if that changes how the cancer cells respond to drugs, the small improvement in animal welfare may render the experiment useless.

Regulations recognize that refinements such as pain medicines and tranquilizers might have a negative impact on data. The AWA requires annual reports of animal use, with a special focus on animals who experience significant unalleviated pain or distress, when scientists are concerned that use of anesthetics, analgesics, or tranquilizers would adversely affect research procedures, outcomes, or data interpretation. Missing from regulations and guidelines is a clear discussion of how unalleviated pain and distress can also affect data outcomes. Pain and stress can affect immune function, social interaction, cancer biology, food consumption, sleep, and other functions, all of which can affect research data. As scientists and IACUCs review whether pain medications and other refinements might disrupt data outcomes, they need to also consider the effects of unalleviated distress and pain.

There are but a small number of model-specific side-by-side comparisons of data outcomes when welfare enhancements are used (e.g., pain medications, stable compatible group housing, enrichments) compared to when they are not. These would be the gold standard to help scientists make determinations and they may sometimes show that welfare refinements affect data in some cases but not in others. When they do, the size of the effect is important; a significant improvement in animal well-being that only minimally affects data outcomes might be worth adopting.

Scientists frequently report quite limited details about the housing, pain management, enrichment, or other refinements in their animals’ lives, making it difficult to establish a standard practice in their field (Würbel, 2007a; Carbone and Austin, 2016). For example, a scientist who believes her peers are not treating pain in a cancer study will believe she cannot use pain medications herself, fearing that pain medications will invalidate her data or make it unpublishable. In reality, it may be the case that her peers simply omit such details from their papers, believing them to be tangential to the data they are presenting. Fuller reporting of animal welfare measures as part of scientists’ Materials and Methods sections of their publication has the potential to both universalize a standard of care for animals, and potentially could improve better research reproducibility (Kilkenny et al., 2010).

As with pain, so is the case with other refinements. Scientists may fear that environmental enrichments such as large complex cages, dietary novelty, social interactions, and opportunities to explore will introduce too much variability in their study populations and will skew their experimental data. And, in truth, animals in such environments may develop differently and behave differently from animals in barren solitary cages. Scientists must consider whether such effects would invalidate their studies, have no strong effect, or possibly improve them (Würbel, 2007b; Li et al., 2013). Scientists must remember that for many well-designed and well-controlled experiments, differing data outcomes are not necessarily ‘truer’; scientists who worry that environmental enrichment, pain medications, or other refinements ‘affect’ animal data may be overlooking the effects of isolation, boredom, pain, lack of thermoregulatory options, and lack of control over one’s life.

19.4 Stress, Distress, and Laboratory Animals

Stress is a generalized response to events, environments, and bodily processes that have the potential to change an animal’s health, safety, or physiological equilibrium (homeostasis). Animals respond to stressors behaviorally, physiologically, or with a combination of the two. Stress is a welfare matter when it rises in severity and/or duration to the level of true emotional distress. Stress can also affect research outcomes, as stressed animals may have profound differences in their biology compared to nonstressed animals. Researchers should bear in mind that even small stressors can affect biology, including behavior, immune function, and metabolism. Minimizing stressors and timing research procedures around stressful events may lead to better data outcomes.

Some stressors pose minimal risk to animals’ well-being and short-term transient stress is inevitable. Even something as seemingly minor as a cage change can induce changes in behavior and activate the classic ‘stress response’ via the hypothalamic–pituitary–adrenal (HPA) axis. However, animals can cope behaviorally if they are able to explore the environment, mark it as they deem necessary, and build their nests. Animals may respond to cold stress behaviorally by building a nest, huddling with other animals, and fluffing up their fur or feathers. Physiologically, they may shiver or activate their brown fat and generate body heat. Thus, behavior and physiology combine to bring the animal to thermal homeostasis.

However, stressors can lead to emotional distress in animals if the stressors exceed the animals’ ability to adjust and cope, and the animal laboratory potentially poses many limitations on animals’ ability to cope. If a cage contains insufficient nesting material, a mouse cannot build a nest for his security, territory, and thermoregulatory needs. Genetically manipulated mice may lack the physical, physiological, or behavioral capacities to cope with some stressors; for example, piloerection is not a physical option for a nude mouse in a cold cage with no fur to fluff up for insulation.

Confronted by multiple serious stressors, animals will prioritize their coping mechanisms and preserve some functions at the expense of others, partitioning their finite physiological resources. Though first described as the ‘fight-or-flight response’, this prioritization of functions is not restricted to confrontations with predators or with bellicose cage mates. If mice in a cold environment with insufficient nesting material shunt their resources toward thermoregulation, this could suppress their immune function and cancer protections. This may decrease their value in some cancer studies as well as decreasing their well-being.

Scientists need to be aware of these stressors, remove the ones they can, limit their duration, and provide the animals with biologically relevant resources to help them cope. A researcher may also choose to accept the effects of stressors and fold that information into their study, which can most easily be accomplished by incorporating them into the experimental design (Richter et al., 2009). This has the added advantage that the investigator need not specifically identify all stressors, but, rather, strive to ensure that they impact all study animals equally. These methods of design may include statistical blocking factors, Latin square designs, housing treatment, and control animals in the same cages, and balancing for housing location.

19.5 Opportunities and Considerations for the Improvement of Well-Being in Laboratory Animals

With these general principles in mind, we next focus on some specific welfare challenges, chosen because of how common they are, how serious they are, or both. This is by no means an exhaustive list of all well-being challenges and opportunities in the laboratory. The ‘Five Freedoms’ that the United Kingdom’s Brambell Report (Brambell, 1965) listed for farm animals in the 1960s also serves as a useful organizing framework for laboratory animals:

As mentioned earlier, most research animals are caged or confined in some way. Important considerations of their confined environment include temperature, light, complexity of the cage, noise, food, and water. The social environment must meet the species’ and individuals’ needs for access to compatible animals but also provide protection from incompatible animals. Many animals want a place to escape and hide from potential predators and aggressors and a way to safely watch, smell, and listen to the world around them.

Beyond review of their housing, we must consider some aspects of the experiments the animals undergo. Pain, contingent or direct, is a crucial factor for laboratory animal well-being. Many welfare concerns straddle the categories of animal housing (or care) and research manipulations (or use). Human caregivers and researchers are an important part of an animal’s world, both for their daily lives and for how people handle them for experiments. Often, research requirements, such as a need to easily restrain animals for examinations, drug administrations, or sample collections, drive particular choices to house animals in smaller cages or in solitary housing.

Researchers have an ethical and regulatory obligation to seek ways to minimize animal suffering. We recommend a birth-to-death inventory checklist of all of the welfare points of concern during an animal’s life in the laboratory, along with all the refinements, analgesics, early euthanasia endpoints, and strategies appropriate for each one. The range of welfare challenges in the animal laboratory require that scientists customize these checklists for their own research program. Table 19.1 illustrates some representative birth-to-death welfare inventories for some common animal research models.

Stay updated, free articles. Join our Telegram channel

Join