The best way to study a human brain is in a living human body, but because there are ethical limits to what can be done experimentally with humans, scientists are left with looking at surrogates, including genetically modified and chimeric nonhuman primates, said Henry Greely, the Deane F. and Kate Edelman Johnson Professor of Law and professor, by courtesy, of genetics at Stanford University. The dilemma, he said, is that as these animals are engineered to more closely resemble humans, the same ethical issues that prevent studies in humans begin to arise. These ethical issues relate to determination of the relationship between benefits and harms, the justification for using nonhuman primates as animal models for research, and the justification for genetically modifying nonhuman primates in ways that may cause pain and distress. Greely added that the public may have unrealistic fears about the humanization of these nonhuman primates and that even scientifically unrealistic fears can have significant consequences for research, and therefore, need attention.
John Morrison, director of the California National Primate Research Center and professor of neurology at University of California, Davis, cited four principles to follow when choosing to work with nonhuman primates: (1) the nonhuman primate should be uniquely well suited to the problem being studied; (2) there should be a critically important health problem closely aligned with the research; (3) the research should be attractive to funding agencies and industry; and (4) there should be a clear route to translation. Joshua Gordon added that studies also should not be conducted in nonhuman primates if they could be done more easily and cheaply in humans, and that nonhuman primate models should not be used for “fishing expeditions,” but for targeted studies where other research in rodent or other models has already provided information on what to seek.
Ethical principles known as the 3Rs (replacement, refinement, and reduction) were articulated 60 years ago (Russell and Burch, 1959) and are now widely applied to animal research throughout the world (Kirk, 2018). They are of particular relevance for research on nonhuman primates (Prescott et al., 2017), especially neuroscience research (Lemon, 2018).
These principles reflect the recognition that nonhuman primates are a unique but limited resource, said Frances Jensen, as well as the fact that like humans, they are sentient, social, and have high cognitive abilities, said Margaret Landi, chief of animal welfare, ethics, and strategy for GlaxoSmithKline.
In terms of harms and benefits, Landi said the 3Rs address only potential harms and direct investigators either to replace an animal with a nonanimal or a lower phylogenetic species, reduce the number of animals, or refine the studies to decrease or eliminate pain or distress. Benefits such as increasing understanding of the disease in question or finding new treatments are generally accepted, she said, adding that benefits potentially could be increased if the translational fidelity of the animal model was increased.
Different Ethical Considerations for Nonhuman Primates Versus Rodents and for Genetically Modified Animals Versus Wild Type
Stefan Treue, director of the German Primate Center in Goettingen, said that regardless of the species, and whether genetically modified or not, animal welfare principles should apply to all animals. However, he said, particular challenges arise with transgenic animals because of the potential harm inflicted by genetic modification and the need to maintain a healthy breeding colony of these purpose-bred disease models. This requires development of new assessment techniques not only in terms of outcome measures, but also to understand the consequences of the genetic manipulation, said Treue.
Some genetically modified animal models are referred to as “humanized” models, although Treue noted that the purpose of genetic modification is usually not to make them phenotypically more like humans, but to build disease models that come closer to the human disease. From an ethics perspective, he said, this is an important and relevant distinction. The main concern, said Treue, is whether the genetic modifications result in unintended (beyond those associated with the disease) consequences that are detrimental for the animal.
Intended consequences are also a concern, said Jeffrey Kahn. If we are concerned about the effects of symptoms in humans, he asked, should we not also be concerned about those effects in the animals that are modeling those symptoms? Moreover, does it matter what disease is being modeled, for example, whether the disease is life threatening or not? He mentioned that similar questions about what counts as sufficiently important research
also arose in the context of gene therapy when the National Institutes of Health’s (NIH’s) recombinant advisory committee considered whether it is appropriate to expose somebody to the risk of gene therapy for a non-life-threatening disease. Regarding humanization, Kahn said questions need to be answered about which capacities matter; how those capacities are assessed as animals are manipulated to act, look, and feel more like humans in their symptomology; and whether humanization in animal models is qualitatively different in nonhuman primates compared with other species.
Xenotransplantation of neural tissue raises other concerns, said Treue, because this involves interfering with exactly the organ that underlies the critical species differences between human and nonhuman primates.
Nonhuman primate research should be held to an especially high standard, said Steven Hyman, Harvard University Distinguished Professor of Stem Cell and Regenerative Biology and director of the Stanley Center for Psychiatric Research. Nonhuman primate studies are ethical only if they are adequately designed and powered, which may require large numbers of animals, he said. Marina Emborg, professor of medical physics at the University of Wisconsin–Madison and director of the Preclinical Parkinson’s Research Program at the Wisconsin National Primate Research Center, suggested that it is too early to use transgenic nonhuman primates for efficacy studies. She agreed with Hyman that unless such studies are properly powered, they are premature, which is the reason that the infrastructure for raising these animals and conducting the studies needs to be increased. Kahn agreed, stating, “It is not ethical to use too few animals and have underpowered research.”
Landi added that lack of robust study designs further compromises the propriety of a study. Unblinded or nonrandomized studies, she said, have a high likelihood of producing a biased outcome that will not translate to people. Guoping Feng added that it is important to realize that monkeys are not human. Researchers need to determine which human characteristics are important to model and recognize that models can only reflect certain aspects of a disease, not the human disease itself, he said. Genetic models may be generated using human mutations—either monogenic or combinations of mutations—that help elucidate the biology of the disease, replicate the circuits implicated in human disease, and thus advance translation of this knowledge to humans.
Emborg reminded workshop participants than an investigator’s idea for the use of an animal model is vetted several times, starting with the call for proposals by federal and private funding agencies, the NIH Blueprint
for Neuroscience Research,1 and peer review. The justification for animal use continues through the review of the vertebrate sections of the proposal and Institutional Animal Care and Use Committee (IACUC) protocols. This multistep process provides guidelines about how to prioritize research and ensure that scarce resources are being allocated in the most efficacious manner. Treue argued that while peer review works for assessing individual proposals, it is ultimately a political or science policy decision where funding lines are established that allow for approaches that break the general mold, including the large, multicenter projects needed for nonhuman primate research.
Feng added that if researchers know of possible approaches that might lead to a treatment and do not pursue them, that also may be unethical. However, the more human-like a model is, and the closer it comes to replicating human disease, the more concerns arise about the ethical implications of making animals sick in order to study them, said William Newsome. He suggested that guidance from those who have already done research with nonhuman primates using transgenic approaches may be helpful. For example, by looking at early studies with SHANK3 mutant animals, investigators may be able to determine whether the mutation made the animals more fragile or difficult to care for, or if they needed special housing or social arrangements.
Regarding the question of whether the care of transgenic and chimeric animals needs to be different, Emborg said that there were many references highlighted throughout the workshop to the use of these types of models that had contributed to human and animal health. In order for those studies to provide valid data, she added, the animals had to receive appropriate care. Emborg said in her MPTP (1-methyl-4-phenyl-1,2,3,6-tetra-hydropyridine) studies, they were prepared to treat the animals with Sinemet if needed, which is the standard of care for patients with Parkinson’s disease (PD). She also used intracarotid rather than systemic MPTP, which produced hemi- rather than bilateral parkinsonism, a milder form of the disease that extends the animals’ lives and allows them to care for themselves.
She also pointed to the dedicated team of veterinarians and animal caretakers that keep close watch on the animals and are prepared to intervene if there are any problems (e.g., if they are not eating or drinking enough, not moving normally, need extra food or warming lamps). They
also go back to the clinic and talk with clinicians who care for people with PD, asking them what their patients are experiencing and what they suggest in terms of caring for the monkeys.
Regulatory Oversight of Nonhuman Primate Research
In the United States, research involving nonhuman primates, whether they are genetically manipulated or not, must adhere to the same rules and regulations that guide all animal research and that are enforced by each institution’s IACUC, according to Landi. She maintained, however, that these rules and regulations should be viewed as the minimum required, and that scientists working with these animals are obligated to go beyond these laws when assessing benefits and harms. While there are no special rules for nonhuman primates, their use raises additional ethical concerns.
Landi outlined principles or commitments that are central to the management of animal facilities, including access to species-appropriate food, water, and housing. If animals have been genetically manipulated, meeting these basic needs may change how they are delivered. For example, modifications to the animals’ environment, feeding, and care may be needed if the manipulations have affected the animal’s ability to chew or swallow, caused mobility or balance problems, or resulted in abnormal behaviors, she said. Ensuring that environments are designed appropriately and that animals receive humane and appropriate care requires a program of veterinarians, behaviorists, and technicians who know how to work not only with the nonhuman primate species, but with how the potential disease or phenotypic change may affect that species, said Landi.
Landi said that looking at the human disease that the animal model is supposed to replicate may help predict what kind of environmental changes will be needed for the animals. However, she added that the species is also important because, for example, macaques are different from marmosets. Although they are social animals, they are also predators who when sick or distressed will hide their symptoms to avoid becoming prey or being rejected by their group.
European animal welfare guidelines overlap those applied in the United States, said Treue. 3R principles are an explicit part of the legal framework in Europe. They were established with the passage in 2010 of a European Union-wide directive on the protection of animals used for scientific purposes. Specific to the use of nonhuman primates is a prohibition against the use of great apes. Every institution is required to have an
animal welfare body roughly comparable to an IACUC, said Treue; however, these committees provide advice and recommendations only. Approval of animal experiments rests with national or regional “competent authorities.”
Hyman suggested that to further align regulatory and ethical guidelines, it might be useful to establish an advisory committee similar to the the consensus committee on the use of chimpanzees in research established by the Institute of Medicine and the National Research Council at the request of NIH in 2010 (IOM and NRC, 2011). The chimpanzee committee established a set of criteria to guide research in an ethical manner, said Kahn, who chaired the committee. Kahn identified several questions that a similar committee focused on nonhuman primates could address in a systematic way:
- What are the relevant reasons to create and use these models (justification)?
- If research on nonhuman primates is justified, what oversight will be needed in order to implement this research?
- What criteria will be used and by whom regarding whether the scientific rationale for a certain study merits approval for that study?
Kahn added that the scientific community should take the lead in developing these guidelines, and that it would be important to invite relevant stakeholders to participate.
Given that the public funds most of the research involving nonhuman primates (other than the use of nonhuman primates for regulatory testing), Treue stressed that researchers have a duty to communicate with the public about the ethical challenges of working with animals in general and more specifically with nonhuman primates. He advocated for a proactive approach, particularly with new technologies in development that may increase the proportion and number of nonhuman primates in research. “We should not view this as a nuisance, but rather an opportunity to do scientific communication and bring the public onboard with why we think this is not only ethically justified, but also scientifically important,” he said.
Gordon added that some members of the public may be concerned about using nonhuman primates for purely fundamental science purposes. However, he argued that a side effect of basic science is learning unexpected things that turn out to be helpful.
Kahn added that improved communication is needed across stakeholder groups to guide research and help refine and articulate the arguments about justification so that any criteria crafted assure that research moves forward in an ethically acceptable way. These communications need to go beyond simplistic justifications, said Treue. It is not enough to say only that scientists want to cure these terrible diseases; further explanations are needed to convey to the public and other stakeholders that research across many labs is needed, he said. Moreover, while no single study will be sufficient, every study if done well can be argued to be necessary, he said.