1

Introduction to Precision Medicine and Animal Models

INTRODUCTION

Precision medicine is focused on the individual and will require the rapid and accurate identification and prioritization of causative factors of disease, said Kent Lloyd, professor and director of the Mouse Biology Program at University of California, Davis, and the co-chair of the National Academies of Sciences, Engineering, and Medicine’s Institute for Laboratory Animal Research (ILAR) Roundtable on Science and Welfare in Laboratory Animal Use. To move forward and accelerate the delivery of the anticipated benefits of precision medicine, he said, developing predictable, reproducible, and reliable animal models will be essential. In order to explore the topic of animal-based research and its relevance to precision medicine, the Roundtable on Science and Welfare in Laboratory Animal Use, with support from the Office of Research Infrastructure Programs at the National Institutes of Health, convened a 2-day workshop on October 5 and 6, 2017.

The idea for the workshop, Lloyd said, stemmed from an editorial that he, Peter Robinson, and Calum MacRae wrote in Science Translational Medicine in 2016 (Lloyd et al., 2016). The editorial, “Animal-based studies will be essential for precision medicine,” included four principles of animal modeling and precision medicine:

• Animal models must reflect -omic variation in patients in order to define downstream functional consequences and discriminate causal from correlative factors at relative efficiency.
• Animal models can effectively link -omic data with environmental, behavioral, and lifestyle information to identify actionable findings.
• Animal models can provide quick and accurate assessment of the scientific validity and clinical utility of gene–environment correlations.
• The incorporation of computational reasoning and semantic mapping efforts will be critical for enabling cross-species phenotype comparisons and maximizing the potential of patient data.

Expanding on these four ideas, this workshop was designed to focus on the development, implementation, and interpretation of model organisms to advance and accelerate the field of precision medicine. In particular, the workshop looked at the extent to which next-generation animal models, designed using patient data and phenotyping platforms targeted to reveal and inform disease mechanisms, will be essential to the successful implementation of precision medicine. The full Statement of Task for the workshop is in Box 1-1.

This workshop was organized under the auspices of the Roundtable on Science and Welfare in Laboratory Animal Use. The purpose of the

Roundtable, said Lloyd, is to facilitate ongoing discussion and information exchange on topics of interest to those involved with the science and welfare of laboratory animals. The Roundtable is made up of thought leaders from academia and the public and private sector, and these workshops are an opportunity for members and experts to engage in discussion in a neutral setting that builds trust and promotes problem solving. The hope is that workshop attendees will be inspired and engaged to share insights and lessons learned with their colleagues and other stakeholders on these topics, said Lloyd, and to encourage cross-fertilization and collaboration between different stakeholders.

ORGANIZATION OF THIS PROCEEDINGS OF A WORKSHOP

The workshop was organized by an independent ad hoc planning committee, in accordance with the procedures of the National Academies. The agenda for the workshop is presented in Appendix A, and the biographical sketches of the planning committee and speakers are presented in Appendixes B and C.

This publication summarizes the presentations and discussions that occurred throughout the workshop. Chapter 1 presents an overview and

history of precision medicine, the regulatory scheme and challenges for regulating precision medicine, issues relating to public opinion and community involvement, and the ethical considerations involved in animal-based research. Chapter 2 presents existing precision medicine initiatives, both in the United States and internationally. Chapter 3 explores the promises and perils of animal modeling: presenters discussed successes, challenges, and unique approaches to animal-based research. Chapter 4 covers issues with reproducibility and predictivity in precision medicine and animal-based research. Chapter 5 looks at in vitro alternatives to animal models, including induced pluripotent stem cells (iPSCs) and microphysiological systems. Presentations that are summarized in Chapter 6 discussed the use of animal models, and the integration of animal and human data, for safety and toxicology assessments. Chapter 7 presents two patient perspectives on precision medicine. Finally, Chapter 8 is a summary of Kent Lloyd’s reflections on the workshop.¹

WHAT IS PRECISION MEDICINE AND HOW DID WE GET HERE?

India Hook-Barnard, director of research strategy at University of California, San Francisco, opened the workshop with a definition of precision medicine: “An emerging approach for disease prevention and treatment that takes into account people’s individual variations in genes, environment, and lifestyle” (NIH, 2015). This definition was first used by the National Institutes of Health (NIH) in the announcement of the Precision Medicine Initiative in 2015, but Hook-Barnard explained that the concept of precision medicine had begun years earlier. In the 1980s, scientific knowledge and technology had reached a point at which it became possible to imagine sequencing the entire human genome. In 1988, the National Research Council published a blueprint for this project, Mapping and Sequencing the Human Genome, that explored technical, legal, and ethical issues as well as funding and organizational needs (NRC, 1988). In 2000, President Bill Clinton announced that the majority of the human genome had been sequenced, and in February 2001, 90% of the human genome sequence was published in the journals Nature and Science. Ten years after this achievement, Francis Collins, director of the National Institutes of Health and former director of the National Human Genome Research Institute, wrote a commentary for Scientific American in which he reflected on the progress

___________________

¹ This Proceedings of a Workshop was prepared by the workshop rapporteurs as a factual summary of what occurred at the workshop. The planning committee’s role was limited to planning and convening the workshop. The views contained in the proceedings are those of individual workshop participants and do not necessarily represent the views of all workshop participants, the planning committee, or the National Academies of Sciences, Engineering, and Medicine.

made and the challenges ahead for precision medicine. In this commentary, Collins repeated Amara’s law: “The First Law of Technology says we invariably overestimate the short-term impact of a truly transformational discovery, while underestimating its longer-term effects” (Collins, 2010).

In 2011, said Hook-Barnard, the National Research Council released the report Toward Precision Medicine, which explored the idea of creating a new taxonomy of disease, based on molecular data and disease mechanisms. The report concluded that “realizing the full promise of precision medicine . . . requires that researchers and health-care providers have access to very large sets of health and disease-related data linked to individual patients” (NRC, 2011, p 7). Hook-Barnard pointed out that the statement of task and the background for that report did not mention precision medicine by name, but that the committee thought about the issues very broadly and determined that a new taxonomy of disease would require these large datasets and linkages to patient data. The report envisioned an “information commons” structure for health data in which each layer contains a specific type of information about patients, for example, data about symptoms, exposures, genome, and microbiome. The information commons approach would be similar to GIS mapping systems, said Hook-Barnard, in that data could be captured, stored, and analyzed using one or more layers (see Figure 1-1). Hook-Barnard added that the report also stressed the importance of participant engagement and health justice. As the field of precision medicine moves forward, it is critically important that representation, inclusion, and engagement be considered and built into the system, said Hook-Barnard.

With a slightly different perspective on precision medicine, David Valle, the Henry J. Knott Professor and director of the McKusick-Nathans Institute of Genetic Medicine at the Johns Hopkins University School of Medicine, argued that “precision medicine” may be an oxymoronic term. The dictionary definition of precise is “exactly or sharply defined,” or “strictly conforming to a pattern, standard or convention”; however, human biology is anything but precise, said Valle. Valle compared two systems—a watch and human biology—and noted that, while both are complex, they differ considerably. A watch is an intentionally designed closed system, with precise yet interchangeable parts. On the contrary, human biology is an imperfect, messy system that developed through evolutionary trial and error. While organs, such as liver and heart, may be partially interchangeable, no two humans are precisely alike. In addition, human biology is an open system, in which human beings are constantly exposed to environmental factors that affect and interact with their biological systems. Valle noted that, in contrast to the old intelligent design theory about human development—in which humans were compared to a watch—the theory of evolution is like a tinkerer without a plan. Valle prefers the term “individualized medicine” rather than “precision medicine,” because it acknowledges the fact that each individual has unique strengths, weaknesses, and needs. However, he noted that the term “precision medicine” has taken hold and he is not likely to change it now.

This said, Valle continued with a description of four principles of precision medicine. First, all disease has both a genetic and an environmental component. He noted that while in some cases—such as Mendelian disorders—the genetic component of the disease is obvious, in reality every disease is the result of an interaction between genetic and environmental factors. Second, everyone has their own disease. That is, two individuals with the same diagnosis may respond quite differently to treatment and may experience the disease in a very different way. Third, disease and risk factors must be considered across the lifetime of an individual. While a precision medicine approach may forestall the onset of a disease when there are known risk factors, this may result in unintended consequences later in the individual’s life. Finally, Valle said that, while change is difficult, the field of medicine can do better and should vigorously pursue precision medicine and what it has to offer patients.

REGULATION OF PRECISION MEDICINE

Robert Califf, Donald F. Fortin MD Professor of Cardiology at Duke University and former commissioner of the U.S. Food and Drug Administration (FDA), added another way of defining precision medicine: the right treatment at the right dose at the right time for the right patient. Califf

noted that, while the term “precision medicine” is fairly new, the FDA has focused on this goal since 1962 with the passage of the Kefauver-Harris Amendments to the Federal Food Drug and Cosmetic Act. These amendments required drug manufacturers, for the first time, to provide evidence that their drugs were effective for the intended use—in other words, that the right patient was getting the right treatment, at the right dose, at the right time.

Califf gave a general overview of the FDA’s mission and responsibilities. The FDA’s mission is “protecting the public health by assuring the safety, efficacy, and security of human and veterinary drugs, biological products, medical devices, our nation’s food supply, cosmetics, and products that emit radiation,” reported Califf (U.S. FDA, 2017). Importantly for precision medicine, Califf added that the FDA is also “responsible for advancing the public health by helping to speed innovations that make medical products more effective, safer, and more affordable and by helping the public get the accurate, science-based information they need to use medical products and foods to maintain and improve their health” (U.S. FDA, 2017). The FDA also regulates labeling of both food and drugs, ensuring that truthful, well-supported, and non-misleading information is available to consumers, he added.

Moving on to the topic of discovery, development, and regulatory approval of precision medicine approaches, Califf noted a number of issues that complicate the process. First, the current drug discovery and development process is years long, and only a very small percentage of initial candidate compounds end up as FDA-approved drugs. There is no reason to believe that the process of developing precision medicine approaches would be any shorter; if anything, it is likely to be longer and more complex. Second, Califf said that there is a dearth of high-quality evidence to support current clinical practice guidelines, even for the average patient (Tricoci et al., 2009). Given this, producing enough high-quality evidence to support a more personalized clinical approach for specific patients will be an enormous challenge. Finally, Califf suggested that changes in the regulatory scheme for new technologies would be needed. Traditionally, tests are regulated one test at a time. For technologies like next-generation sequencing (essentially 3 billion tests all done at one time), Califf believes there needs to be an extensive evidence base underpinning this technology. Otherwise, the regulation of next-generation sequencing still won’t be accomplished years from now. He added that for other advances, such as gene editing, it is critical to have high-quality animal models that can help us understand the potential for off-target effects and long-term effects.

Despite these challenges, Califf stated that he believes that there is great promise in precision medicine. Califf said that in order to realize the vision of precision medicine, it is critical that the research community

develops an ecosystem for nonhuman or animal studies that accounts for this complexity, develops reproducible predictive models, and can respond rapidly to the need for co-clinical mechanistic studies. At the same time, the challenge for the FDA is developing a system that ensures safety in the face of all these scientific advancements, while not inhibiting the opportunity for innovation.

Califf told workshop participants about the Project Baseline Study, a collaboration among Verily, the Duke University School of Medicine, Stanford University, Google, and the American Heart Association. Project Baseline is aimed at developing gold standard data, tools and technologies to provide a holistic view of human health and more efficiently and effectively conduct clinical research. Participants in the study undergo comprehensive assessments, including the following:

• Genomics, proteomics, transcriptomics, metabolomics, microbiome, etc.
• Eye exam, audiometry
• Personal and family medical history
• Imaging chest X-ray, coronary computed tomography, echocardiography, etc.
• Cognitive and physical tests
• Blood, urine, stool, saliva tests, microbiome

Some of the information gathered by the Baseline Study is collected through passive sensors, such as watches that continuously record physiological and environmental data, sleep sensors, and an app that works as an interface for self-reported as well as passive data collection. Using these technologies is a huge advance, said Califf, because previous longitudinal studies were limited by the need to get participants into the clinic repeatedly for data collection.

Another precision medicine initiative that is under way is PCORnet, also known as the National Patient-Centered Clinical Research Network, said Califf. PCORnet, which is funded by the Patient-Centered Outcomes Research Institute, is a large, highly representative, national patient-centered clinical research network that seeks to conduct large-scale, relevant clinical research, with the end goal of enabling people to make informed healthcare decisions. The data in PCORnet come from more than 122 million patients, collected by 34 different health systems. This large-scale data network enables researchers to perform passive observational studies as well as prospective studies, Califf noted.

Califf concluded with his “regulatory dream” of what would be needed in order for precision medicine to realize its promise of delivering the right treatment at the right dose at the right time for the right patient:

• Integrative nonclinical models that provide reliably probabilistic estimates of translation into human biology
• Quality management systems that enable assurance about data provenance, analytical appropriateness, and reproducibility
• Rapid, less expensive, and scalable models to test mechanisms when unexpected results occur
• Ecosystem engagement in rapid learning, given the new capacity to acquire, store, and analyze digital information
• Deliberate risk taking to advance beyond regulatory inertia to move the field forward

ETHICAL CONSIDERATIONS OF ANIMAL-BASED RESEARCH

Elizabeth Heitman, professor in the Program in Ethics in Science and Medicine at the University of Texas Southwestern Medical Center, opened her presentation by saying how important it was that this workshop was being held, noting that it was the perfect time to discuss the ethical issues surrounding precision medicine and animal-based research. She explained that, if ethical discussions are held too early in the process of developing a new scientific advancement, there is not enough known about it to predict its likely future manifestations, potential applications, and possible effects. However, if these discussions are held too late, the technology may already be so diffused into clinical care that it is too late for ethical considerations to have any impact through professional norms and standards, formal policy and regulation, or public opinion.

Heitman noted that workshops held by the National Academies are often designed to explore emerging technologies with an eye to anticipatory governance. In other words, the workshops are designed to engage stakeholders in strategic, policy-oriented discussions about risk reduction and management, using predictive processes to identify key decision points and the necessary steps for managing the development, application, and effects of new technologies. She expressed her hope that this workshop would contribute significantly to future governance of animal-based research for precision medicine.

Consideration of ethics and values is essential for anticipatory governance, said Heitman. Professional ethical standards of practice shape how new scientific developments unfold and inform how they are applied. The public’s diverse values support—or oppose—scientific developments and new technologies, often from distinctly different perspectives than those of the scientific community. Heitman noted that, while the public has been engaged in the Precision Medicine Initiative, there has been little ethical discussion around the role of animal-based research for precision medicine.

Ethicists, like most other participants in anticipatory governance, are

challenged by the need to predict how a technology will develop and be applied. Typically, clear ethical standards in science have been articulated retrospectively, such as after a crisis or scandal or after dramatically new capability challenges the status quo. Ethicists must rely on others’ reports about new scientific developments to identify their potential ethical impact and make recommendations about their ethical use, Heitman said, thus discussion of animal models in precision medicine is still too new to have generated a related ethical discourse. Heitman pointed out that, in fact, the first outline about animal models in precision medicine was published in 2016, in the article quoted by Kent Lloyd in his introductory remarks (see p. 2; Lloyd et al., 2016).

To begin the discussion about the ethics of animal-based research for precision medicine, Heitman walked workshop participants through the three “Rs” of animal-based research (Russell and Burch, 1959).

With these ethical principles for animal-based research in mind, Heitman, referring to the definition at the National Library of Medicine website,² identified four elements of precision medicine research that would affect the practice and implementation of the three Rs:

1. The reliability, sharing, and use of animal data
2. Modeling genetic effects on environmental exposures
3. Modeling genetic effects on lifestyle and behavior exposures
4. Modeling genetic effects over a lifetime

___________________

² See https://ghr.nlm.nih.gov/primer/precisionmedicine/definition (accessed March 5, 2018).

Data

Precision medicine will require large databases for storing, analyzing, and sharing data on both humans and animals. Heitman said that collaborative databases can help support the goals of reduction and replacement, because researchers can answer research questions using already existing data, rather than carrying out new animal experiments. She noted that the ethical norms and best practices for data sharing are still being defined and adopted, and that funding is needed in order to establish collaborative databases and to facilitate sharing and communication. Furthermore, data sharing and coordination remain challenging, both in terms of creating systems that can “talk” to one another, and in terms of patient confidentiality and institutional policies about data sharing. She listed several existing collaborative databases for animal-based research:

• WormBase
• FlyBase
• Mount Sinai’s Center for Personalized Cancer Therapeutics’ database of fruit fly tumor avatars and effective chemotherapy
• Zebrafish International Resource Center (ZFIN)
• The Knockout Mouse Project
• International Mouse Phenotyping Consortium
• International Mouse Strain Resource
• Immuno Polymorphism Database/Non-Human Primates Database

Gene–Environment

One type of precision medicine research that is likely to utilize animal-based models is research on how gene variants interact with exposure to environmental factors, such as pollution. Heitman noted that this type of research has the potential to conflict with existing standards for safe animal housing, particularly when the research focuses on studying complex environmental exposures such as multiple pollutants or sick building syndrome (situations in which people who live or work in buildings experience adverse health effects that seem to be linked to spending time in the building, but there is no specific illness or cause identified). She also noted that research on environmental factors has the potential to run counter to the goal of reduction: when there is already a public health solution for an issue (e.g., reduce industrial pollution), it may be unethical to conduct research on animals rather than pursuing the public health solution. Finally, Heitman observed that these types of studies may be problematic not just for animals but for humans as well. She suggested that research that is aimed at ameliorating the impact of environmental factors—for

example, developing a drug to counter an industrial pollutant’s effects on heart health—risks doing injustice to humans by medicalizing what would otherwise be an environmental issue.

Gene–Lifestyle

Another potential area of animal-based research for precision medicine is the study of how genes interact with behaviors or lifestyle choices. In designing this type of research, Heitman said, researchers will need to carefully distinguish between behavior that is volitional—for example, cigarette smoking—and factors that are not volitional—for example, exposure to wood smoke. She said that this type of research may require the use of more sentient animals that can express volition rather than be passive participants in an environmental hazard, and noted that this would conflict with the goal of replacement.

Longitudinal Data

In order to fully understand how different genes lead to different outcomes over a lifetime, researchers may need to collect data on animals for their entire natural lifespan. Heitman said that this may conflict with the goal of refinement. Collecting data about prolonged, chronic, and terminal illnesses may conflict with the goal of using refined, early end points in animal studies; research animals would normally be humanely euthanized before they were subjected to diseases such as advanced cancer. In addition, using animal models for studies on diseases linked to long-term environmental exposures will require exposing animals to harmful and potentially stressful settings for long periods of time, which conflicts with the ethical standards of housing and care currently employed.

Heitman concluded that, in addition to these four examples, there will likely be old, new, and potentially unexpected ethical issues that arise with the use of animal-based research for precision medicine. Stakeholders—including researchers, journal editors, Institutional Review Boards, funding agencies, and patients—need to be on the lookout for these ethical issues and be ready to engage the public in a discussion of how to design high-quality research that is consistent with the three Rs of animal-based research.

BIOETHICS, PUBLIC OPINION, AND COMMUNITY ADVISORY BOARDS

Richard Sharp, professor of biomedical ethics and medicine and director of the Biomedical Ethics Research Program at the Mayo Clinic,

said that there is a “perfect storm” around the issue of animal research that involves declining public support for animal-based research, increasing concerns about certain kinds of research—particularly around genetic modification—and increasing political activism about these issues. Sharp said that the percentage of Americans who believe it is “morally acceptable” to perform medical testing on animals has dropped from 65% in 2001 to 51% in 2017, representing a significant slide (Gallup News, 2017). At the same time, emerging technologies, such as CRISPR/Cas-9, are generating public concern about the ability to fundamentally alter the genomes of organisms; these concerns fall into several general categories. First, there is public concern about catastrophic global consequences, for example, from genetically modified mosquitos. Second, there are concerns about animal experimentation that would make animals more like human beings for the purpose of investigating their biology, said Sharp. Third, there are experiments that raise concerns about the integrity of species, for example, the human neuron mouse that was proposed at Stanford over a decade ago.

In the context of animal modeling for precision medicine, Sharp foresees potential for controversy in several areas. First, the public’s concerns over animal research often center on the question of whether it is reasonable to inflict suffering on animals for what may be an uncertain benefit to human beings. Certain precision medicine initiatives in particular may spark this concern, for example, the mouse “hospitals” in which humans and mice undergo co-clinical trials to inform treatment of the human (see Chapter 3). While this model may be promising for identifying effective therapies for patients, it may be a disturbing image for many people in the public, said Sharp. The fact that strains of animals would be engineered solely in order to benefit an individual patient could generate significant controversy and opposition. Second, animal research raises the question of how many mice should be sacrificed in order to save a human life. As precision medicine research advances, it is likely that animal models will play a significant role, and this increase in the need for animals in research may draw opposition. While these general areas of concern about animals are not novel, the rise of precision medicine may put increased focus on animal research.

Given this potential for controversy, Sharp had suggestions for moving forward. First, he said that we are often in a state of discourse without dialogue, in which people talk at, but not with, each other. The challenge, said Sharp, is to move beyond this and to really engage in conversation with people in order to understand where these concerns are coming from. Second, Sharp said that there is a lack of trust in the scientific community that can result in opposition to some forms of research. He quoted E.F. Einsiedel, who said: “While public concerns are often dismissed as naïve or misguided, the public use of dystopian images also reflects their lack of

trust in the scientific community and their skepticism in the capacity of government to regulate in the public interest” (Einsiedel, 2005, p. 154). Sharp said that more effective engagement and dialogue with “opponents” can help bridge this divide and increase trust in the scientific community. Finally, Sharp encouraged scientists to learn from historical mistakes such as the story of Henrietta Lacks and the Havasupai controversy.

The key to improving dialogue and engendering trust is community engagement, said Sharp; one way to accomplish this is through Community Advisory Boards (CABs). These CABs, which exist across the country and comprise people from various parts of the local community, can be extremely effective for providing researchers with insights about whether and how members of the community might object to the work they are doing. The Mayo Clinic, said Sharp, has three CABs that meet a handful of times throughout the year and provide input to researchers with regard to governance structures, procedures, and donor engagement. In addition to providing ongoing feedback, a CAB can also work with researchers to engage the community on specific topics or areas of concern. For example, Sharp said that there was an issue concerning partnerships with industry and the commercialization of biobank samples, and it was necessary to do a deep dive with the members of the CAB and really discuss the issues at hand.

As an example of how CABs can operate, Sharp gave a few details about one of the Mayo CABs. Advisors on the CAB are drawn from the southeastern Minnesota area and the Biobank participant population and serve a term of 2-3 years. They attend five or six meetings per year, and receive $75 compensation for each meeting. Advisors receive regular communications between meetings and are expected to interact respectfully with one another, recognizing the plurality of views among the advisors. Sharp said that the Mayo Clinic makes it clear to the CAB advisors that their role is to serve as representatives of their communities, but not as regulators. He said that the advisors are not asked to approve or disapprove research projects, and he believes it is important to be clear with CAB members what authority is being granted. The Mayo Clinic CAB has advised the Biobank leadership on topics, including the following:

• Informed consent process
• Communications about biobank research
• Underrepresented minorities
• Return of genetic findings
• Partnerships with industry
• Commercialization of biobank samples
• Use of new technologies, such as whole-genome sequencing

Sharp stressed that the goal of CABs is not necessarily to come to consensus, nor to argue opponents into submission. Rather, the value of CABs is to clarify moral differences and identify areas where there seems to be some irreconcilable, substantive, and moral disagreement in play, and perhaps to find agreement on certain moral positions for which there may be agreement. This process requires recognizing marginalized voices and recognizing divergent moral views. Ultimately, “by engaging opposing views we may not be able to reach consensus but we may be able to find some type of compromise position for which no one is happy but everyone is at least satisfied,” said Sharp. The value in this type of dialogue, said Sharp, is that it might help “avoid political entrenchment and the kind of moral extremism that’s become all too common in political debates in the U.S.” Sharp added that CABs are not the only mechanism by which these difficult goals can be met, and that there are more formal and less formal ways to structure these types of conversations.

Given the potential for controversy in precision medicine research, Sharp said that the scientific community should expect there to be public concerns and should engage the public in conversations in order to build trust and increase the transparency of the work. He cautioned, however, that, while this engagement is valuable and necessary, it is unlikely to reduce these debates to a place where all stakeholders come together and rally in support of the same goals. Despite the challenges, Sharp said that discussions about precision medicine research should be expanded in order to include the debates about animal research. Although this may be an uncomfortable thing to do, Sharp said that discussions about animal research will be an essential part of the conversation that needs to take place.