Michelle Mello from Stanford University, a member of the Symposium Planning Committee, introduced the theme of the session. Key Findings 2 and 3 of the National Science Advisory Board for Biosecurity (NSABB) Working Paper (NSABB, 2015a: 3-4) addressed the adequacy of the policy frameworks in the United States to provide oversight of gain-of-function (GOF) studies of concern. Finding 2 indicates that the frameworks are effective overall, yet Finding 3 suggests that their adequacy for managing the risks associated with GOF research may vary, depending on which pathogen is being studied.
She commented that there seemed to be plenty to discuss about where the policy frameworks may and may not be adequate or optimal for addressing these risks. To that end, the speakers were asked to reflect, depending on their institutions, on the issues facing federal agencies in administering this regulatory framework as well as some of the strengths and weaknesses in the current policy framework and opportunities for optimizing oversight of this area of research.
Gerald Epstein from the Department of Homeland Security reviewed the scope of the NSABB proposal in terms of who is covered, which pathogens and activities are covered, and what is required. In terms of the existing policy context, he wanted to differentiate between those that are in effect by force of law, and therefore affect all researchers in the United States, as opposed to those that are, for example, a condition of government funding, which would directly affect only the recipients of that
funding. There may be indirect effects in other areas, but a funding hook would only directly affect recipients of U.S. government funding.
The first part of the existing regulatory context that affects everyone in the United States is the laws in place to prohibit biological weapons development.1 This statute is the mechanism by which the United States implements the Biological and Toxin Weapons Convention, an international treaty which prohibits development or acquisition of biological weapons. Unfortunately, the law does not contain definitions of prohibited types of activity or agents, so he thought the level of subjective judgment involved in proving a violation makes prosecution difficult.
Partly for that reason, the Select Agent Regulations were developed and expanded through a series of statutes.2 This is a comprehensive set of safety and security requirements governing any use of certain listed pathogens. Of the GOF pathogens, three of them—1918 flu, highly pathogenic avian influenza, and severe acute respiratory syndrome (SARS)—fall under these regulations. The Middle East respiratory syndrome (MERS) does not.
Under the Select Agent Regulations, institutions have to be registered, researchers and staff have to be vetted by the government, and the institution has to have permission to use those agents. There are requirements for safety and security and for incident response and reporting associated with use of these pathogens. And in this case, the government does not have to prove intent. If one is found with one of these agents and has not registered with the government, it is a violation of a law. When that law was passed, it was also recognized that there are legitimate and important reasons why these agents need to be used in research. This is why there is a process by which research institutions and people can become vetted and approved to work with these agents. But it does provide a barrier for people who are not within that scheme.
A third area of legislation that binds everyone in the United States is export controls. These affect the export of certain listed pathogens from the United States or the communication of certain nonpublic, proprietary information that could, for example, include information about how to develop a particular strain of a pathogen if that were not published in the open literature. Information that is published in the course of fundamen-
1 The primary statute to implement U.S. obligations under the Biological and Toxin Weapons Convention is the Biological Weapons Anti-Terrorism Act of 1989 (Public Law 101-298, May 22, 1990).
2 The Select Agent program was created by the Antiterrorism and Effective Death Penalty Act of 1996 (Public Law 104-132, April 24, 1996). Following the attacks of September 11, 2001, and anthrax mailings, the program was expanded by the USA PATRIOT Act of 2001 (Public Law 107-56, October 26, 2001) and the Public Health Security and Bioterrorism Preparedness and Response Act, known as the Bioterrorism Act of 2002 (Public Law 107-188, June 12, 2002).
tal research is not affected by export controls, but there is a set of statutes and regulations that could have some bearing on the ability to do and disseminate biological research.
Because they had already been discussed, Dr. Epstein touched only on the policies that are attached to government funding. This includes the federal and institutional policies for oversight of dual use research of concern (DURC), which among the GOF pathogens covers only 1918 flu and H5N1 highly pathogenic avian influenza (White House, 2012, 2015b).
He commented that another framework developed by the Department of Health and Human Services (HHS) for certain H5N1 and H7N9 strains was very similar to the structure of the NSABB’s recommendations (HHS, 2013). Lawrence Kerr would describe the framework during his presentation.
Dr. Epstein cited Biosafety in Microbiological and Biomedical Laboratories (CDC and NIH, 2007) and Guidelines for Research Involving Recombinant or Synthetic Nucleic Acid Molecules (NIH, 2013), extensive biosafety and biosecurity guidance that the Centers for Disease Control and Prevention (CDC) and the National Institutes of Health (NIH) developed for use by anyone doing biological research. It is obligatory as a condition of NIH funding for institutions to follow these processes. This set of best practices enables research using potentially dangerous pathogens to be done safely both for those working in the laboratory and for those in the community. Beyond the formal requirements, this guidance is used widely around the United States, not only for government-funded work, and indeed around the world. Even a policy that nominally has only the force of its ties to government funding can have much greater influence.
Finally, with the caveat that he was not a lawyer, Dr. Epstein cited the issues of the liability that any institution working with potentially hazardous substances could face. Any entity working on something that could pose a risk to its workers, to the neighborhood, or to the environment has to do so in recognition that if there were an accident that causes damage in the community they could be held financially liable. This includes not only harm to the institutions or employees but also harm to the general public. And the extent to which an institution could be held liable may depend on the degree to which there is a regulatory structure in place and whether the institution had been complying with those regulations.
Dr. Epstein commented that any additional development of policy related to GOF research would have to be embedded in the already existing frameworks, and the question of whether these existing procedures would have to be modified to fit the new one or whether they would sit on top would have to be determined as the policy process went ahead.
Dr. Kerr from HHS provided an overview of HHS framework for research with certain highly pathogenic avian influenza viruses (HHS,
2012). Dr. Kerr noted that during the research life cycle, there are points at which biosecurity concerns could be addressed, but it is too difficult—and too damaging to the research enterprise—to do this at the publication stage. Therefore, HHS focused on the research proposal stage as part of the funding award process. He noted that within the existing HHS GOF policy, the focus is on studies that could produce an agent with increased pathogenicity or transmissibility via respiratory droplets. In such cases, an extra level of review is required. The results of the review determine whether a proposal goes on to departmental level review. Seven criteria are taken into account during these reviews (see Box 3-1).
The department-level review provides multidisciplinary expertise—including public health, scientific, security, intelligence, countermeasures, and preparedness and response—from a number of agencies to evaluate these proposals. The department-level review will also identify any additional risk mitigation measures that should be required, and it will determine whether a given proposal is acceptable for HHS funding. For proposals that are deemed acceptable, the funding agency within HHS will make the final funding decision. Dr. Kerr indicated that only a small number of research proposals had undergone a departmental review, but he commented that the results reflected the full spectrum of what one might expect from a review process if it is working well. Some proposals received full approval by the committee and were recommended for funding to the funding agency director. There were also research proposals
that were received in which individual experiments were rejected by the committee and it was recommended to the funding agency that those not be funded.
Richard Frothingham from Duke University provided an overview of the review process for dual use research instigated by their Institutional Biosafety Committee (IBC), which has been reviewing research for dualuse potential since 2003. Its experiences were recounted in an article in Science in 2007 (Davidson et al., 2007). The committee determined that most projects with significant dual use potential were GOF studies, and as a result, it added seven questions to its recombinant DNA registration form in 2005. It also undertook specific training for IBC members on dual use research. The IBC has examined all research including recombinant DNA, select agents, and all research under BSL3 conditions as well as other research upon request.
The Duke IBC does not use a specific dual use definition or threshold but has identified relevant research through the NIH study section or program officer, by the principal investigator’s (PI’s) answers on the recombinant DNA registration form, or by members of the IBC during the course of its research reviews. Specific examples of GOF identified by the Duke IBC were provided, including cytokine expression by Ectromelia; virulence factors in uropathogenic E. coli; adaptation of dengue virus for growth in Drosophila cell lines; and HIV infectious molecular clone pseudotyped with vesicular stomatitis virus-G (VSV-G) for initial entry into renal cells. The Duke IBC had learned a number of lessons from having reviewed this research, including:
- GOF studies were encountered regularly as part of the broad category of dual use research, but the IBC had yet to encounter GOF studies of concern.
- PIs have had challenges with the concept of dual use research; it was possible to reach consensus on the dual use potential of most biomedical research, but not on specific categories (e.g., DURC).
- Focusing on specific risk mitigation strategies rather than whether a particular experiment was DURC did enable the IBC to reach consensus, and no GOF research proposals have been rejected.
- The Duke IBC had received external expert advice on some studies as part of the review process and this had been helpful.
Dr. Frothingham noted that the comparative scarcity of events involving the misuse of research to cause harm makes it difficult to measure the benefits from dual use reviews. He did highlight their value in building public trust in responsible science. He suggested that the early review of GOF research might reduce wasted effort by scientists and improve peer
review and funding outcomes. He concluded by providing a number of perspectives on the NSABB recommendations (see Box 3-2).
Philip Potter from St. Jude Children’s Research Hospital introduced the hospital’s work on influenza, including its status as one of the National Institute of Allergy and Infectious Diseases’ Centers of Excellence for Influenza Research and Surveillance and a World Health Organization Collaborating Center for Studies on the Ecology of Influenza in Animals. Dr. Potter noted that, as a result of their work with influenza viruses, St. Jude is likely to be affected by decisions over the oversight of GOF studies of concern. He highlighted the existence of a specific DURC committee that consists of both scientists and nonscientists. In their system, the PI is responsible for presenting the risks and benefits of the proposed studies. To assist the committee, St. Jude has developed internal guidance on what they should consider. This included ensuring that no GOF virus is resistant to antiviral agents, that suitable vaccines are available, and that advice about the challenges of evaluating risks and benefits in “gray” areas, such as research altering host range and or tropism, is available. The committee has also embraced the ferret as the gold standard for biological testing, requiring its use in all relevant experiments.
The DURC committee has also subjected all experiments involving H7N9 influenza virus to the same scrutiny as the 15 agents covered by DURC requirements. Dr. Potter noted that the DURC committee does
not publish minutes. He also provided a number of perspectives on the NSABB draft recommendations and broader policy frameworks (see Box 3-3).
The discussion that followed addressed local adaptation of the research covered in assessment of DURC or GOF studies of concern. J. Patrick Fitch from the Battelle National Biodefense Institute, a member of the NSABB, speaking in his personal capacity, raised the question of who would be responsible if something went wrong—the scientist or the committee? He also commented that his institutional committee had an experience similar to Duke’s. In that situation, a focus on developing appropriate risk mitigation plans for relevant research, rather than on identifying a specific experiment as “DURC,” had proved to be a much more productive approach to achieving the same goal.
Following a suggestion by planning committee member Barry Bloom from Harvard University, speaking in his personal capacity, participants also explored whether there was a need for separate IBCs, DURC committees, and possibly GOF studies of concern committees. Some participants felt that it might be possible to combine committees, especially if there was access to the additional expertise that might be required for new roles. Allison Mistry from Gryphon Scientific also proposed updating the DURC requirements to reflect GOF studies of concern and to
avoid creating a separate definition and policy oversight process for GOF research. Christopher Park from the Department of State stressed the importance of the scope of GOF studies of concern in covering both biosafety and biosecurity issues, which set it apart from the DURC process. Richard Frothingham expressed a concern that the DURC institutional process was considerably more cumbersome than the normal IBC process, and he would be reluctant to see them combined.
Dr. Frothingham highlighted the importance of the independence of review committees and their ability to access external expertise. He and Philip Potter discussed the advantages, for example, of including local public health officials in the membership of the IBCs, which both Duke and St. Jude do. Both also have regular contact with the Federal Bureau of Investigation’s local Weapons of Mass Destruction coordinator. The importance of clear definitions was stressed by Diane DiEuliis from the National Defense University, while others, such as Mr. Park, highlighted cases where overly detailed definitions undermined the intended aim of the measure. There was a discussion about whether it was better to limit the scope of research likely to be captured under these definitions or, as proposed by Mr. Park, to have a fast-track process for removing research not deemed relevant during the review process. The unique nature of each research proposal was stressed by Dr. DiEuliis, as was the need to consider each proposal in context.
Several participants noted the importance of exploring alternative approaches to GOF studies of concern whenever possible, and the panelists discussed several specific examples of this happening. The value of broader expertise and nonspecialists in identifying alternative research approaches was noted in this regard by Drs. Frothingham and Potter.
Michelle Mello and several of the panelists felt that public trust was an important metric for assessing the efficacy of regimes for DURC and GOF studies of concern. Other participants suggested that reviews of DURC and GOF studies of concern were sensitive and should not be publicly available. Some participants argued that transparency was important and that relevant records should be made available. Both Dr. Frothingham and Gerald Epstein commented on the difficulties posed by the competing goals of protecting potentially sensitive information and ensuring transparency as part of gaining public trust. There were suggestions by Mr. Park, for example, that such information might be made available but not widely distributed.
The discussion also identified a number of tools that might strengthen future efforts. Gregory Koblentz of George Mason University highlighted the importance of learning from past experience. He and other participants called for mechanisms to capture lessons learned in a more systematic fashion. Professor Koblentz also called for additional help for PIs
to understand the underlying concerns that drive assessments of DURC and GOF studies of concern. He also proposed more support to assist regulators in understanding what is possible at the laboratory level and to enable public understanding of the research.
There was also some discussion of the proposal from Silja Vöneky from the University of Freiburg to require laboratories to take out insurance against the risks of GOF studies of concern. Dr. Epstein saw the utility of this proposal—the Department of Homeland Security requires the laboratories it funds for Select Agent research to have insurance. He felt that it would be a useful approach for improving good behavior, although he saw challenges for insurance companies in developing accurate actuarial calculations on these risks.
The definition of GOF studies of concern contained in the NSABB WG’s Draft Working Paper was revisited. The third criterion for defining GOF studies of concern was once again the most discussed by some participants, such as planning committee member Dr. Bloom from Harvard University, speaking in his personal capacity, finding no difference between the transmissibility criteria and the one connected to innate or acquired resistance to public health interventions. In connection to the third criterion, Dr. Bloom also raised issues of justice and equity around access to drugs in many parts of the world.
Philip Dormitzer from the Pfizer Vaccine Research and Development Unit, a member of the Symposium Planning Committee, introduced the session as a continuation of the earlier plenary session on the U.S. policy landscape. This session would present the perspectives of several different key stakeholders, including regulatory agencies and the vaccine industry.
Michael Callahan from the Massachusetts General Hospital and Harvard Medical School opened by highlighting that U.S. efforts to balance the risks and benefits of GOF could be altered for use in other contexts and adapted to the needs of different countries. Dr. Callahan stressed the interconnected nature of the research and development enterprise in the life sciences and for biotechnology, asserting that “the world is flat for bio-innovation.” As an example, Dr. Callahan noted that more viral pathogens have been sequenced in China in 4 months than have ever been sequenced in the United States and Europe.
A major theme of Dr. Callahan’s remarks was that market-driven and beneficent GOF research is already happening around the world, all of which is outside of the U.S. and European policy and regulatory frameworks. One of his main points was that U.S. and European vaccine
production does not always take cultural and other factors into account. For example, he noted that Western vaccines will only be used in Indonesia if they conform to requirements that make them halal. Another relevant example discussed was the production in Asia of effective and inexpensive H5N1 poultry vaccines. The life span of a chicken in Asia is about 6 months, so Western vaccines costing $7 per dose are not going to be used when Asian-produced vaccines costing pennies per dose are available. He also noted that countries in the group of 112 Non-Aligned Nations may refuse to share pathogen gene sequences with U.S. scientists because “they’ve been ripped off.” He suggested that the U.S. government needs to protect “our international collaborators from R01-funded investigators who seek to do nothing more than get a virus, go home, and write their big Nature paper.” Dr. Callahan concluded with a series of recommendations for aligning domestic and foreign policies relevant to GOF (see Box 3-4).
Robert Fisher from the Food and Drug Administration (FDA) introduced the FDA’s main mission: to ensure that medical products and associated technologies are safe and effective. Dr. Fisher discussed a variety of regulations relevant to the evaluation of products and the implementation of regulatory mechanisms. He highlighted a number of approaches used by the FDA, including randomized clinical trials, surrogate endpoints, and animal efficacy data. He stressed that, regardless of the approach taken, the FDA relies on data for its decision making.
Dr. Fisher noted that the GOF framework focuses on specific agents of concern, or particular pathways of concern, potentially impacting
FDA-relevant research. However, the narrowing of focus to GOF studies of concern reduces the potential impact considerably. He provided examples of where GOF studies of concern might impact the work of the FDA, including the production of vaccine seeds from molecular clones, or adapting vaccine candidates to grow in cell-based systems rather than egg-based systems. He noted the potential for GOF concerns to impede rapid, large-scale production of vaccines to meet seasonal and emergency needs.
Lessons from the FDA’s regulatory experience were also provided: for example, the importance of early and sustained engagement with stakeholders. Dr. Fisher stressed the value of ensuring sufficient flexibility in regulatory regimes and for their implementation.
Jonathan Moreno from the University of Pennsylvania framed his comments as both a bioethicist and a patient—or a “consumer” of the public health benefits of GOF research. Dr. Moreno identified a number of questions that he thought needed to be addressed when considering policy options for dealing with GOF research, including the impact of terminology and the response to the term GOF; the implications of mutants—both natural mutations and escape mutants; the potential for a generalized GOF policy being too broad to implement effectively; the adequacy of safety records for quantifying the risks of laboratory accidents; current levels of accident reporting; the need to address basic research and vaccine development activities where more acceptable or safer alternatives do not exist; determining the realities of the relationship between GOF methods and vaccine development; and the role of basic science during a health emergency.
Dr. Moreno identified five areas where he thought there was consensus relating to the controversy over research and the policy options to address it:
- Much regulation fails to hit the mark for this field and could needlessly delay vaccine development
- Some regulation is needed, for both biosafety and biosecurity
- Biocontainment does not have a perfect record
- Risk mitigation often only requires some imagination
- Sometimes there are acceptable alternatives to GOF studies of concern, even if they are not the best option
He also thought there might be agreement that GOF data alone cannot predict emergence of a pandemic (genotypes to phenotypes), but perhaps this is getting better; the long-term potential of pre-pandemic strain selection could be “transformative” in new vaccine development; humans are vulnerable to certain natural strains that could be targeted for research,
such as bat SARS-like coronavirus strains; and animal model development for SARS and MERS should be permitted.
He highlighted opportunities for public deliberation as to whether all three of the characteristics for GOF studies of concern, as proposed by the NSABB, are needed for an experiment to warrant additional oversight, or whether the production of a pathogen anticipated to possess two of the characteristics would be sufficient. Dr. Moreno highlighted the need to build on best practices when developing capacities to review GOF studies of concern. He reviewed the composition and mission of the Wisconsin Bioterrorism Task Force and the Stanford University benchside ethics consultations as examples to be considered. Dr. Moreno presented a potential model for institutional bodies for the operational review of GOF studies of concern: Risk–Benefit Assessment Teams, or R-BATs (see Box 3-5).
Ethan Settembre from Seqirus provided an overview of the global influenza system that addresses variability in influenza viruses to develop and deliver candidate vaccine viruses. Dr. Settembre described an example of how the system works in practice, detailing vaccine generation over a 4-5 month period in response to the H1N1 influenza pandemic in 2009. He noted that while vaccines can be produced increasingly quickly, pandemics
emerge even more rapidly, necessitating further research and development. Dr. Settembre highlighted a need to further speed up production of vaccines in response to both pandemic and seasonal influenza events. He discussed a synthetic process for generating vaccine candidates using attenuated backbones and available hemagglutinin and neuraminidase sequences that was developed with the J. Craig Venter Institute and Synthetic Genomics Vaccines, Inc.. The process allows the generation of synthetic influenza viruses that are attenuated, but would allow for speed, accuracy, and high yield. He noted that this is one of the ways to make vaccine viruses to address immediate important medical needs in a short period of time to get ahead of the wave of infection. This approach had been used to produce an H7N9 vaccine candidate in 2013.
The discussion that followed included both an interchange among the panellists and questions and comments from the audience. Among other topics, the discussion returned to considering different ways of defining GOF studies of concern. Philip Dormitzer pointed out that not all GOF research involves GOF studies of concern, and therefore not all the research needs to be overseen by any additional policy frameworks. Following his comments, panelists and participants discussed the subjective nature of determining what is (and what is not) of concern. Michael Callahan argued that any definition for GOF studies of concern needs to be general, adaptive, and culturally appropriate for foreign scientific communities. Participants noted that specialist terminology might not translate well into other languages and settings: for example, Dr. Callahan noted long-standing issues around the meaning of “biosecurity.”
Participants also reexamined the importance of ensuring that the process to consider policy options, as well as any new frameworks it produces, cover both the public and the private sector. The importance of adequate containment for GOF studies of concern was another reoccurring theme during the discussions. The issue of enforcement was also revisited. Possible unintended consequences for new policy frameworks to oversee GOF studies of concern were discussed by Dr. Callahan, and possibilities of a negative impact on vaccine production were considered. The risks of regulatory uncertainty were also addressed, with some participants arguing that regulatory burdens are more acceptable when the “what” and the “why” are clear. Issues around the harmonization of domestic oversight regimes, such as those for DURC and GOF studies of concern, were also highlighted by Robert Fisher. Some participants called for the development of a more overarching framework to deal with risks and benefits from life sciences research.
The international implications of determining thresholds of concern or acceptable risk were considered, as well as international perceptions about why the United States might be concerned about this research. Dr. Callahan suggested that it was important to understand the nature and motivation of relevant international stakeholders to improve the dialogue on GOF research. To this end, he noted the importance of strengthening research collaborations, in particular working more closely with partners inside their countries. Gregory Koblentz asked whether it was time to move beyond stovepiped concepts of “biosafety” and “biosecurity” to adopt a more holistic concept of “biorisk management.” Dr. Fisher responded that, from a regulator’s viewpoint, to the extent that such an approach could reduce uncertainty, it could be helpful.
There was an exploration of the impact of over-regulating GOF research for countermeasure development. Dr. Dormitzer pointed out that, because one of the factors for identifying GOF studies of concern is the absence of effective countermeasures, limiting research that could provide such measures could be counterproductive. There was also consideration of the opportunity costs of not doing research, especially in justifying potential barriers to developing countermeasures. Issues around intellectual property were also explored, with Dr. Callahan discussing barriers for the development of countermeasures, or barriers to the conduct of science internationally. There was a call to change the incentives for countermeasure development—to produce more players, more stakeholders, and therefore more solutions. To this end, Dr. Callahan recommended that greater attention be paid to foreign industry, as an increasing number of products and self-sufficient markets were being developed.
Challenges in disease surveillance were discussed. While some participants suggested that knowledge produced by GOF research could be useful for detecting emerging pathogens, others noted the lack of current surveillance capacity. Current shortcomings in data sharing and capacity for disease surveillance can also distort risks from disease and the impact of public health measures, according to Dr. Callahan. He also noted international concerns that disease surveillance data and capabilities are being used for nonpublic health purposes.