New pathogen research rules: Gain of function, loss of clarity

In December 2017, after six years of debate and discussion, the United States government closed a chapter—though perhaps not the book—on one of the most controversial experiments in the annals of dual-use research: the creation of an H5N1 avian influenza virus that was transmissible through the air between mammals. That is, the Health and Human Services Department has finally issued new rules governing how it will decide whether to fund similar experiments in the future. While these new rules embody a reasonable set of principles for assessing the risks and benefits of such research, the review process could also be strengthened in several ways to ensure that it is comprehensive and rigorous.

The rise of “gain of function.” In 2011, Dutch researcher Ron Fouchier and American virologist Yoshihiro Kawaoka sparked a major controversy when, separately, they submitted manuscripts to Science and Nature in which they described how to generate strains of H5N1 avian influenza that were transmissible between mammals. This research on mammalian airborne transmissible H5N1, or matH5N1, was published only after the National Science Advisory Board on Biosecurity and the World Health Organization engaged in lengthy and contentious review processes that focused on whether the research should be made public.

The experiments—which not only demonstrated that mammalian transmission of the virus was possible but also provided information on how to construct such a virus—triggered broad concern about the safety and security of so-called “gain-of-function” experiments. “Gain-of-function” experiments were defined by the US government as experiments that resulted in the creation of pathogens with enhanced virulence or transmissibility, or both. The introduction of the term “gain of function” immediately caused confusion. From a scientific perspective, “gain of function” can refer to a wide range of natural or artificially induced biological mutations, some of which could potentially create a more dangerous pathogen and others that do not. Because the controversy over matH5N1 was framed in terms of “gain of function,” and not according to existing terminology and concepts for describing and assessing dual-use research that had been developed over the previous 10 years, the life sciences and biosecurity communities had to endure a long, unproductive debate about how to define this new category of worrisome research. Carving out “gain of function” as somehow distinct or separate from dual-use research has muddied debate and has had a long-lasting impact on policy.

The controversy over experiments that could make H5N1 transmissible between mammals triggered a flurry of new policies on dual-use research in the United States. In 2012 and 2014, the US government issued new regulations describing oversight of dual-use research of concern. The term “dual-use research of concern” was defined as comprising seven types of experiments conducted on one of 15 pathogens or toxins that, because they posed the highest risk of deliberate misuse, were designated as Tier 1 agents. Despite these new rules, concern about gain-of-function research was renewed when the Centers for Disease Control and Prevention and the National Institutes of Health, in July 2014, experienced a trio of biosafety failures involving variola virus (the causative agent of smallpox), Bacillus anthracis (the bacterium that causes anthrax), and avian influenza. None of these incidents caused any human illnesses, but they occurred over a short time and at elite biomedical research institutions; these circumstances heightened concerns about the safe conduct of research into pathogens with enhanced virulence or transmissibility.

In response to the incidents, in October 2014 the White House issued a moratorium on funding new gain-of-function studies on influenza, SARS, and MERS; asked scientists engaged in such research to voluntarily halt their experiments; and announced that it would initiate a “deliberative process” to develop a new policy on dual-use gain-of-function research. The deliberative process, which was led by the National Science Advisory Board on Biosecurity, and which lasted 18 months, consisted of five meetings of the board; the commissioning of both a technical report analyzing the risks and benefits of gain-of-function research and an ethical study of such research; and two workshops, conducted by the National Academies of Science, to solicit the input of stakeholders. In May 2016, the advisory board issued its recommendations for oversight of gain-of-function experiments. (The recommendations introduced yet another confusing term: “gain-of-function research of concern.”)

In January 2017, the deliberative process reached its culmination with the issuance of guidance by the White House Office of Science and Technology Policy on research involving enhanced pathogens of pandemic concern. The shorthand for this guidance is “the P3CO Framework,” where—in government-speak—”P3CO” stands for “potential pandemic pathogen care and oversight.” The problematic terms “gain of function” and “gain-of-function research of concern” were thankfully replaced by language built around a better term—“enhanced.”

In its guidance, the Office of Science and Technology Policy recommended that federal agencies adopt new mechanisms to govern the creation, transfer, and use of potential pandemic pathogens with enhanced virulence, transmissibility, or both. According to the guidance, agencies adopting such policies might be able to resume funding the research that the White House had suspended in October 2014.

In December 2017, the Health and Human Services Department—home of the National Institutes of Health, the world’s largest funder of biomedical research—issued its policy on the oversight of research involving enhanced potential pandemic pathogens. This new policy—called the “Framework for guiding funding decisions about proposed research involving enhanced potential pandemic pathogens”—is broadly similar to, but not identical to, the corresponding guidance from the Office of Science and Technology Policy, which itself drew heavily from the advisory board’s report. The department, at the same time it issued its policy, also announced that, under the new oversight system recommended by the Office of Science and Technology Policy, it was resuming funding of research into enhanced potential pandemic pathogens.

We, the authors, harbor concerns about adequate oversight of potentially dangerous research, and the framework incorporates several elements that address those concerns. The framework is thorough. It does a good job of laying out the principles and processes through which the Health and Human Services Department will make funding decisions regarding research that involves enhanced potential pandemic pathogens. The framework’s approach to dual-use research of concern is not based on lists of experiments or on specific pathogens, but instead takes a risk-based approach that focuses on the attributes of modified organisms. While the identity of starting organisms is central to existing oversight policy for dual-use research of concern, the framework emphasizes the importance of organisms’ properties once the experiment is over. This more comprehensive approach to dual-use research is a welcome change. Some elements of the new framework, however, remain worrisome.

Too narrow. The framework’s scope is not broad enough. The framework applies only to research funded by the Health and Human Services Department, and primarily the National Institutes of Health. Research funded by other federal agencies, or by the private sector, is not subject to this review process. This narrow scope contrasts with rules that the United States issued in 2012 regarding oversight of dual-use research, which applied government-wide. Key provisions in the Health and Human Services framework contain wording almost identical to that in the guidance issued by the Office of Science and Technology Policy, so there does not appear to be a compelling reason that every federal agency should need its own set of rules to oversee research involving potential pandemic pathogens. Yet other agencies do need to oversee such research. The Agriculture Department has identified research on transmission of avian influenza between species, including mammals, as a key gap in understanding the epidemiology of this virus, and one that needs to be filled. Potential biodefense research, especially research involving threat characterization, could be conducted or funded by the Defense Department, the Homeland Security Department, or the intelligence community, and such research could be relevant as well.

Likewise, considering current growth in the bioeconomy, increasing commercialization of synthetic biology and genome editing, and the increasing role of the private sector in funding basic research, the exemption from oversight of privately funded life sciences research is a large and growing loophole. The recent synthesis of horsepox virus by Canadian scientists, with funding from a US biotech company, illustrates how privately funded research can stray, intentionally or inadvertently, into the realm of dual-use research. The National Science Advisory Board on Biosecurity recommended that oversight of gain-of-function research be applied to all researchers, regardless of their source of funding. While this approach has not yet been implemented, the guidance issued by the Office of Science and Technology Policy held out the prospect of revisiting oversight of potential pandemic pathogens to determine if such oversight could be extended to all life sciences research, regardless of source of funding.

Terminology and definitions. The framework’s definitions of certain terms leave a lot to be desired. For example, the framework defines a potential pandemic pathogen as a pathogen that has both of the following properties:

1. It is likely highly transmissible and likely capable of wide and uncontrollable spread in human populations; and

2. It is likely highly virulent and likely to cause significant morbidity and/or mortality in humans

This definition of a human potential pandemic pathogen has some flaws. The word “likely” is too strong, as it implies a high probability or a high level of confidence in the estimated probability. “Possibly” would be a better qualifier—it implies some probability, but does not set the bar too high. Words such as “highly” and “significant” might be too strong as well.

Assessing the risk posed by research with potential pandemic pathogens requires consideration of several factors. These include the likelihood that an experiment will generate an organism with enhanced virulence, transmissibility, or both; the likelihood that this virus could escape the laboratory; and the consequences of such a release. An estimated 60 percent case fatality rate is associated with wild-type H5N1 avian influenza. So even if the likelihood of an experiment generating a strain of this virus with enhanced transmissibility among mammals were judged as possible but not likely, grave concerns about the virus’s potential to seed a pandemic if it escaped from a laboratory would remain high.

As another example, consider a highly pathogenic avian influenza strain that has been modified in the laboratory to bind to isolated human lung-cells in culture—a first step toward infection in humans via airborne transmission. If these modified viruses have not been tested for airborne transmissibility in ferrets, the standard animal model for studying human influenza airborne transmission, would they qualify as being “likely highly transmissible” among humans and therefore subject to review under the framework? The framework should not be interpreted to apply only to reviewing in vivo work with enhanced potential pandemic pathogens. In vitro research with an enhanced virus that could possibly be transmissible among humans should still be subject to review.

The framework does a better job defining what an enhanced potential pandemic pathogen is—or, in the older vernacular, a gain-of-function potential pandemic pathogen. An enhanced potential pandemic pathogen is defined as “a [potential pandemic pathogen] resulting from the enhancement of the transmissibility and/or virulence of a pathogen.” This definition is straightforward. The “and/or” language here is important, as the literature is rife with research studies in which the virulence of avian or human influenza viruses has been intentionally or accidently increased in ferrets, mice, and other mammals. Such research, based on the definition cited just above, should be subject to review under the framework as well.

Review process. The review process created by the new rules is a mixed bag. One positive development is that entities such as institutional biosafety committees, which review research to ensure that it adheres to relevant biosafety and dual-use research rules, have been removed from the process of deciding if a proposed project should be referred to the Health and Human Services Department for review. Institutional biosafety committees may not have the expertise necessary to review certain types of research proposals, and they could be pressured by investigators within their institutions to deem proposed research not subject to review by Health and Human Services.

Instead, the funding agency holds the primary responsibility for implementing these new rules. First, the funding agency “conduct[s] standard scientific merit review” of a proposal. Next, proposed research that is “reasonably anticipated” to create, transfer, or use enhanced potential pandemic pathogens is referred to departmental-level review. The department-level review is conducted by a multidisciplinary group including experts with experience in scientific research, biosafety, biosecurity, medical countermeasures, law, ethics, public health preparedness and response, biodefense, select agent regulations, and public health policy. The review group can also contain voting and non-voting members from Health and Human Services or other federal agencies. The funding agency must “consider the recommendations resulting from the departmental-level review” to make a decision on funding.

Multidisciplinary review is an important aspect of dual-use research oversight, and its inclusion in the new rules, along with the wide-ranging expertise suggested for the proposed department-level review committee, is to be applauded. But multidisciplinary review committees should include some ad hoc members from outside the government—otherwise, the details of reviews may never be transparent to outsiders.

An important question remains about where the review group will be based. It is unclear where the departmental review would be managed. The guidance by the Office of Science and Technology Policy, in order to reduce the risk of conflict of interest, encouraged agencies to vest responsibility for oversight of enhanced potential pandemic pathogens in offices that are not part of an agency that is proposing to fund such work. This recommendation was a direct result of a controversy surrounding H5N1—the advisory board got into trouble for questioning the wisdom of experiments that had been funded by the National Institutes of Health, the board’s parent agency. Subsequently, most of the board’s members were replaced, the charter of the group was narrowed, and no meetings were held for two years. Any review of research into enhanced potential pandemic pathogens needs to be organized and conducted by a group sufficiently independent from the funding agencies that have an interest in the research under review. The Office of the Assistant Secretary for Preparedness and Response in the Health and Human Services Department is one logical choice.

Another positive aspect of the framework is that it requires the department-level committee to consider the risks posed by research that involves any of the following outcomes:

• Enhancing the harmful consequences of a pathogen
• Disrupting immunity or the effectiveness of an immunization against the pathogen without clinical or agricultural justification
• Conferring to the pathogen resistance to clinically or agriculturally useful prophylactic or therapeutic interventions against that pathogen, or facilitating the pathogen’s ability to evade detection methodologies
• Increasing the pathogen’s stability, transmissibility, or ability to disseminate
• Altering the host range or tropism of the pathogen
• Enhancing the susceptibility of a host population to the pathogen
• Generating or reconstituting an eradicated or extinct pathogen

These seven categories of experiments are based on a list first put forward 14 years ago in a National Academy of Sciences report titled “Biotechnology research in an age of terrorism” (also called the Fink Report). This is the same list of experiments included in the 2012 and 2014 US government policies for dual-use research of concern—but those policies are only applied to research with 15 Tier 1 pathogens and toxins. The inclusion of these seven categories of dual-use experiments in the framework usefully expands, beyond increases in virulence and transmissibility, the range of potential risks that will be subject to review by the Health and Human Services Department. Still, since the new department policy only applies to enhanced potential pandemic pathogens, it represents only incremental progress toward the Fink Report’s recommended goal of reviewing all research in the life sciences that engages in such experiments.

New criteria for risks and benefits. Assessing the risks and benefits of research into enhanced potential pandemic pathogens is inherently problematic. During the deliberative process, participants never achieved agreement on whether potential benefits outweighed risks for matH5N1. Many still believe that the potential benefits of that research do not justify the potential risk. How will the risks and benefits of research into enhanced potential pandemic pathogens be measured and balanced? What criteria will the Health and Human Services Department use to guide its review of research proposals and its funding decisions? What is the threshold or standard of evidence that will be used to make these judgments? Most of the criteria discussed in the framework are standard in existing policy regarding dual-use research of concern, and will generate the same intensity and diversity of opinion that were witnessed throughout the debate on H5N1 and during the deliberative process.

The framework contains two notable items that are not standard in existing policies for dual-use research of concern. First, review of proposed research with an enhanced potential pandemic pathogen will include an assessment of whether there are “no feasible, equally efficacious alternative methods to address the same question in a manner that poses less risk than does the proposed approach.” During the deliberative process, some observers pointed out that alternative methods have been published that can determine the mutations required to make H5N1 avian influenza transmissible through air between mammals—and that these methods would not generate new strains of enhanced live viruses. Researchers should always employ low-risk or no-risk alternatives first, before resorting to live virus—especially one with the potential to cause a global pandemic if it escaped from a lab. Therefore, it is good to see this principle of responsible research enshrined in the framework.

The second addition to the criteria for guiding funding decisions by the Health and Human Services Department is laudable, but it may prove more difficult to operationalize. The final criterion in the Health and Human Services review process, which is based directly on recommendations by the advisory board and the Office of Science and Technology Policy, is to determine whether or not research is “ethically justifiable.” The department-level review committee, in its review of the ethical aspects of a proposed experiment, is encouraged to consider to what extent the experiment represents non-maleficence, beneficence, justice, respect for persons, scientific freedom, and responsible stewardship. Previous guidance for dual-use research of concern was focused strictly on scientific criteria for assessing the risks and benefits of dual-use research. In a 2015 paper commissioned, as part of the deliberative process, by the advisory board on the ethics of dual-use research, Monash University professor Michael Selgelid observed, “Like risk-benefit assessment, ethics involves inevitable uncertainty.” The Health and Human Services Department, facing an extra dimension of uncertainty in the review process, will be forced to grapple with a new set of issues which it might find itself unprepared to address adequately.

Transparency. The criteria used to judge which experiments involving enhanced potential pandemic pathogens warrant review by the Health and Human Services Department—and how the risks, benefits, and ethical aspects of such experiments are measured and weighed—are ambiguous enough to provide departmental reviewers wide latitude in their funding decisions. The process and outcomes must be transparent in order to demonstrate that the process is conducted in good faith and that policy is implemented appropriately. The framework, though it recognizes the importance of transparency for maintaining public trust in science, does not go far enough in actually providing the requisite level of transparency.

The guidance from the Office of Science and Technology Policy called on agencies “to the maximum extent possible” to “provide transparency to the public regarding funded projects involving the creation, transfer, or use of enhanced [potential pandemic pathogens].” According to the framework, the Health and Human Services Department “will periodically ask the National Science Advisory Board for Biosecurity to review the process described herein.” This approach to transparency is neither timely nor tied to specific reviews of proposed or funded projects. Instead, the policy provides only for occasional transparency about the process itself, not the results of that process. In addition, nothing guarantees that the advisory board review will be made available to the public. Furthermore, since the advisory board is advisory only, it has no ability to force Health and Human Services to revise its process if it finds the process lacking, or to revisit or reverse a funding decision. Meanwhile, the advisory board itself recommended the establishment of an independent advisory committee to evaluate the implementation of the new oversight policy. Until such an accountability mechanism is established, the Health and Human Services Department should provide annual reports on the implementation of its policy on potential pandemic pathogens, which would be reviewed by the advisory board and interested stakeholders. It is also incumbent upon the Office of Science and Technology Policy to follow up on the commitment it made in its guidance to review the implementation of departmental policies on review of potential pandemic pathogens one year after their adoption and determine if any revisions are necessary.

International considerations. The final weakness in the framework is that it only applies to research conducted within the United States. Previous research has demonstrated that research with enhanced potential pandemic pathogens is occurring in labs around the world. Ultimately, the release of an enhanced potential pandemic pathogen from a laboratory is an international issue, as pandemics know no boundaries. With modifications, the framework could serve as a starting point for international discussion about oversight and regulation of research into enhanced potential pandemic pathogens. In December 2017, state parties to the Biological and Toxin Weapons Convention—the international treaty that outlaws the development and possession of biological weapons—agreed to establish a number of expert working groups to examine key issues related to the treaty. One of the working groups will be dedicated to examining developments in science and technology related to the treaty, including biological risk assessment and management. Another working group will focus on strengthening national implementation, including measures to prevent the development of biological weapons and the transparency of dual-use research. Both of these working groups could provide suitable venues for discussing new US policies on the oversight of research into enhanced potential pandemic pathogens and for learning how other countries approach this issue.

The December 2017 adoption of the framework by the Health and Human Services Department is a milestone in the long-running debate over how best to provide oversight of life sciences research to maximize benefits and minimize risks. While this policy was the culmination of several years of effort by a diverse group of stakeholders, it is by no means the final word on the subject. The Office of Science and Technology Policy is supposed to lead a review of policies on potential pandemic pathogens one year after their adoption by departments. This article has provided an initial assessment of the areas that deserve further scrutiny—and possibly revision—in order to strengthen oversight of this important field of research.