Otana Jakpor is a senior double majoring in Global Health and Biological Sciences, with plans to matriculate into medical school in the fall. When she is not happily harmonizing with Overflow A Cappella or dutifully delving into rigorous readings, she enjoys injecting as much alliteration as practically possible into her prose (frequently to her friends’ frustration!).
In September 2011, Dutch virologist Dr. Ron Fouchier announced at an influenza conference that his lab had engineered a mutant form of the a/H5N1 virus (avian flu) to be transmissible between mammals (Cohen, 2012, p. 1155). Around the same time, Dr. Yoshihiro Kawaoka’s lab at the University of Wisconsin-Madison independently achieved the same result. To date, there has been a 54% fatality rate in confirmed cases of avian flu. However, there have only been about 600 confirmed cases in the first place, because avian flu has generally only spread to humans directly from infected birds (World Health Organization/Global Influenza Program [WHO/GIP], 2012). Publication of these labs’ work was temporarily halted for fear that evildoers would replicate their methods and create a global pandemic, and some have even contended that these studies should not have been conducted at all. Scientists counter that such research is vital so that societies can prepare ourselves for future pandemics, and I agree. However, although the security risks associated with this type of research may be exaggerated, they do exist. In light of these considerations, I believe that a new international mechanism should be developed for handling biological research with serious security implications.
Initial reactions to the creation of aerosol-transmissible a/H5N1 were mixed, but a predominant theme was fear. Fouchier’s announcement was met with panic from the media, as exemplified by a Daily Mail headline that proclaimed, “Man-made flu virus with potential to wipe out many millions if it ever escaped is created in research lab” (Daily Mail Reporter, 2011). Alhough the United States government did not join in this overtly hysterical response, it was alarmed as well. The National Science Advisory Board for Biosecurity (NSABB) reviewed Fouchier’s and Kawaoka’s manuscripts, which had been submitted to Science and Nature, respectively, and advised that the papers not be published in their entirety, with procedural details being omitted (Novossiolova, Minehata, & Dando, 2012, p. 41). Specifically, the NSABB was concerned that “publishing the methodological and other details of this work could potentially enable replication of experiments that had enhanced transmissibility of a/H5N1 influenza… by those who might wish to do harm” (National Institutes of Health [NIH], 2011). Their recommendation was non-binding, since the NSABB is limited to an advisory role (U.S. Secretary of Health & Human Services, 2012). Both Nature and Science deferred to the NSABB’s counsel, and the authors of the manuscripts (as well as other scientists conducting research on H5N1) “voluntarily suspend[ed] certain research on the H5N1 virus for 60 days,” to allow time for international dialogue on the issue (NIH, 2011). That being said, editors have not always been so acquiescent. Regarding a controversial publication on the synthesis of the Spanish Influenza virus, the then editor of Science declared, “So would I… have published the paper even if the NSABB had voted otherwise? Absolutely — unless they had it classified” (Kennedy, 2005, p. 195). If publishers take this attitude again in the future, the NSABB will not be in a position to stop them.
“So would I… have published the paper even if the NSABB had voted otherwise? Absolutely — unless they had it classified”
The NSABB’s primary concern was dual use: that nefarious individuals, such as terrorists, would exploit the research for purposes other than the intended goal of developing tools to protect public health. One member of the NSABB bluntly explained, “We don’t want to give bad guys a road map on how to make bad bugs really bad” (Enserink, 2011). However, the researchers have countered that terrorists are unlikely to have the expertise and resources necessary to engineer human-to-human transmissible a/H5N1, even if given explicit instructions. Moreover, the methods used by Fouchier’s and Kawaoka’s labs are not novel (Saey, 2012, p. 20). As Fouchier and his colleagues stated matter-of-factly, “Individuals with bad intentions do not need to read the details in our manuscript because the methods for creating similar viruses have already been published widely” (Herfst, Osterhaus, & Fouchier, 2012, p. 1629). Although dual use is a reasonable concern, it is too late for censorship of publications like this to be effective, even apart from the ethical problems with redacting research.
Much of the concerns over dual use have been centered on terrorism. Either by genuine forgetfulness or deliberate omission — and I suspect it is the latter — many published opinions on this issue have left out the possibility of a/H5N1’s being used as a biological weapon by countries, not terrorists, in times of war. Due to the anthrax attacks and other recent terrorist attacks, it is understandable that bioterrorism rather than biological warfare would be a leading concern for the United States at this point in time. Nevertheless, as Dr. Michael J. Selgelid, Director of a World Health Organization (WHO) Collaborating Centre for Bioethics at the Australian National University, has noted, “Considering the size of microbes compared with nuclear warheads, proliferation is an especially serious concern in the context of biological weapons” (Selgelid, 2012, p. 38). Terrorists may be unlikely to have the necessary technology to synthesize the a/H5N1 virus, but some countries would certainly have the required resources. It is therefore evident that the current system for handling the publication of research that may pose a biosecurity threat, which varies by country, is inadequate for addressing ethical problems along these lines.
In defense of his study, Kawaoka did not discount the possible risks associated with a/H5N1 research, but he argued that not researching the virus would be even riskier. Kawaoka countered, “Because a/H5N1 mutations that confer transmissibility in mammals may emerge in nature, I believe that it would be irresponsible not to study the underlying mechanisms” (2012, p. 155). It is quite possible that a lethal a/H5N1 pandemic may arise in nature anyway, so, as Fouchier and his colleagues observed, “[c]ensoring the manuscripts on a/H5N1 virus transmission will… only create a false sense of security” (Herfst et al., 2012, p. 1630). The only way to evaluate the likelihood of deadly mutations in viruses is to conduct research like that of Kawaoka and Fouchier. Furthermore, the only way to make use of research is to share the findings with others who can follow up with either more studies or a plan of action. Unless research on lethal viruses is conducted and shared via the appropriate channels, the world cannot prepare for the next global pandemic.
In defense of his study, Kawaoka did not discount the possible risks associated with a/H5N1 research, but he argued that not researching the virus would be even riskier.
Although much of the debate surrounding this research has dealt with the issue of publication, some critics question whether these studies should have been conducted in the first place. A New York Times editorial remarked, “In this case, it looks like the research should never have been undertaken because the potential harm is so catastrophic and the potential benefits from studying the virus so speculative” (“An engineered doomsday,” 2012). Some worry that a laboratory accident could result in the public’s being accidentally exposed to a/H5N1; as this editorial put it, “the consequences, should the virus escape, are too devastating to risk” (“An engineered doomsday,” 2012). The danger of human error in hazardous biological research is indeed a legitimate concern, but standards and protocols do exist. Theoretically, at least, these expertly-devised precautions ought to reduce risk to a manageable level.
Fouchier and his colleagues have insisted that “[r]esearch on class 3 pathogens, including transmissible HPAI a/H5N1 virus, can be done safely under enhanced BSL3 [biosafety level 3] conditions by well-trained laboratory professionals, using strictly defined biosecurity and biosafety regulations” (Herfst et al., 2012, p. 1629). They went on to note that even if one individual were to be exposed to the virus as a result of inevitable human error, H5 vaccines, antiviral drugs, and quarantine are options for preventing further spread of the disease. However, although they made a convincing case that “the risks for the public and the environment can be reduced to nearly zero,” they provided little reason to believe that the risks will be minimized in every case (Herfst et al., 2012, p. 1629). Perhaps their lab genuinely was responsible in handling the virus safely, but other labs may not be. Although human error is not sufficient reason to halt a/H5N1 research entirely, it does highlight the need for a large-scale system of safety standards for dangerous biological research.
This need for oversight and implementation of safety and security standards was addressed at a meeting at the WHO in February 2012, where experts recommended that the biosecurity and biosafety requirements needed for this type of research be reviewed (Herfst et al., 2012, p. 1630). They also concluded that “hiding key details from publications does not serve science or public health and that the confidential sharing of data classified as ‘dual use of concern’ is impossible within a relevant time frame” (Herfst et al., 2012, p. 1630). The NSABB thus reconvened and recommended that Kawaoka’s revised paper be published in full, and that Fouchier’s full manuscript be published as long as he were to add “a number of scientific clarifications” and remove “additional information that would enable the construction of an H5N1 virus that is both highly pathogenic and transmissible between animals through the air” (National Science Advisory Board for Biosecurity [NSABB], 2012).
Although human error is not sufficient reason to halt a/H5N1 research entirely, it does highlight the need for a large-scale system of safety standards for dangerous biological research.
Part of the reason behind this apparent change of heart was that the U.S. Government had just released the United States Government Policy for Oversight of Life Sciences Dual Use Research of Concern. Alhough this new policy allows for classification of research, it does not provide a mechanism for sharing the data on a need-to-know basis; classification would therefore interfere with scientific collaboration between the U.S. other countries (Novossiolova et al., 2012, p. 44; Branswell, 2012). The NSABB thus decided that redaction was an “unworkable” option, and reluctantly concluded that it would be better to publish the studies in full than not publish them at all (Novossiolova et al., 2012, p. 44; Branswell, 2012).
However, although Fouchier and Kawaoka eventually both had their papers published, the moratorium on a/H5N1 research that was originally intended for 60 days was lengthened after the WHO meeting. Participants at the meeting came to a consensus on the following point:
[T]he biosafety and biosecurity conditions under which further research is conducted on the laboratory-modified H5N1 viruses should be fully addressed by relevant authorities…. [T]he current moratorium on research to enhance the transmissibility of H5N1 influenza viruses and the further research on the laboratory-modified viruses should continue until the conditions have been determined. (WHO, 2012)
While reviews and clarifications of appropriate biosafety and biosecurity conditions have been carried out in some countries, it has yet to be done in others (Fouchier, García-Sastre, & Kawaoka, 2012, pp. 1-2). As a result, the moratorium has continued to this present time, and it seems that it will persist indefinitely. This is unfortunate, because influenza is a perpetually urgent topic of research, due to the high rate of virus mutation in nature.
There have been international attempts to address the overlap between security and the life sciences in the past. In 1972, the Convention on the Prohibition of the Development, Production and Stockpiling of Bacteriological (Biological) and Toxin Weapons and on their Destruction (or the BTWC) was created as a supplement to the Geneva Convention. However, although it is considered bad form to break an international convention, there is no system in place to monitor compliance to the BTWC (Dunworth, Mathews, & McCormack, 2006, p. 93). Consequently, according to a study conducted in 2003, only 47% of states parties had legislation in force implementing the BTWC (Woodward, 2003, p. 12). As it stands, the BTWC is not a sufficient instrument to hold governments accountable in the area of research oversight.
The BTWC’s fundamental problem is its lack of a forceful implementation mechanism. Dr. Marc Lipsitch and Dr. Barry R. Bloom of the Harvard School of Public Health have proposed that the WHO assemble an international group “to set guidelines and oversee international laboratories working on such pathogens [like a/H5N1]” (2012, p. 3). They went on to note, “Getting agreement will be difficult, and enforcement will be very challenging; but when reducing risk, an imperfect system is much better than none” (Lipsitch & Bloom, 2012, p. 3). Perhaps this group should set standards not only for international laboratories, but for domestic laboratories that handle pathogens posing great risk to global health. It is important to have a certain level of consistency, both within countries and across international lines.
The BTWC’s fundamental problem is its lack of a forceful implementation mechanism.
So far, the WHO has only formed two legally binding instruments, but it certainly has the capacity to negotiate more (Silberschmidt, 2012, p. 17). The Constitution of the WHO states that one of its functions is “to propose conventions, agreements, and regulations, and make recommendations with respect to international health matters and to perform such duties as may be assigned thereby to the Organization and are consistent with its objective” (WHO, 2006, p. 2). Moreover, the WHO Health Assembly has the “authority to adopt regulations concerning… sanitary and quarantine requirements and other procedures designed to prevent the international spread of disease” (WHO, 2006, p. 7). Practical enforcement of an instrument to regulate research on pathogens like a/H5N1 would still be a significant challenge, but worth pursuing under the auspices of such an established and respected organization as the WHO.
Ideally, unacceptably hazardous studies should be prevented before any research is actually conducted, such as during the funding or institutional approval phases, to avoid surprise censorship controversies. If, despite these precautions, a journal has cause to believe that a submitted manuscript might be too risky to be published, it should forward it to an appropriate committee under the WHO for review. In this situation, depending on the sensitivity of the findings, data from select labs might be shared only with each other and with public health institutions on a need-to-know basis. These public health institutions would then take further steps such as calling for more research to address a particular question raised by the initial study, or development of tools to combat the health threat.
Ideally, unacceptably hazardous studies should be prevented before any research is actually conducted, such as during the funding or institutional approval phases, to avoid surprise censorship controversies.
Although the controversy surrounding the publication of Fouchier and Kawaoka’s papers has been tensely resolved, it has highlighted the pressing need for a mechanism to regulate how sensitive research is conducted and shared on an international level. The WHO should immediately begin the process of developing a treaty that creates an international group to provide oversight for studies on a/H5N1 and similar pathogens Ideally, the WHO will begin this process by forming a working group by the end of this year to draft elements of this treaty. There should be a representative distribution of key stakeholders and experts within the working group, to avoid bias. The ongoing moratorium on certain a/H5N1 research is the perfect opportunity for reform of the way this type of research is conducted and shared. Even though developing an effective, appropriate mechanism for this will be challenging, it is absolutely vital in order to both further scientific knowledge and protect public health.
Update as of March 6, 2013: Since the time when this essay was written, several important developments have taken place. First, researchers ended their voluntary moratorium on February 1, 2013, a little over a year after it began. They stated, “We fully acknowledge that this research… is not without risks. However, because the risk exists in nature that an H5N1 virus capable of transmission in mammals may emerge, the benefits of this work outweigh the risks” (Fouchier et al., 2013). That said, they also asserted that the moratorium should continue in countries that had not yet established policy for research on H5N1 transmission, such as the United States (Fouchier et al., 2013). The moratorium ended in the United States later that month, when the U.S. Department of Health and Human Services and the NIH announced a new framework for dealing with H5N1 research (Greenfieldboyce, 2013). Meanwhile, on February 21, 2013, the White House Office of Science and Technology Policy released for review a proposed new policy addressing “dual use research of concern” (Jones, 2013). Finally, last week (February 26-28, 2013) the WHO held an “Informal Consultation on Dual-Use Research of Concern,” with a report to be posted on their website in due course (WHO, 2013).
For a bibliography, please contact Scribe at scribe.usc@gmail.com
Leave a Reply