Scientific Publication and Global Security

Legitimate biological research that can be deliberately misused to cause a biological threat to public health, national security, or both, termed dual use research (DUR), poses critical issues for the scientific, medical, and public health communities. To carry out terrorist events successfully, a terrorist or a rogue state must have the necessary information, equipment, and expertise. Publishing information that enables or facilitates terrorism anywhere in the world is of grave concern. The utility of published scientific information to terrorists depends on its quality and level of detail. Publishing “low-tech” as well as “high-tech” research results can present risks.¹ Although the biodefense and preparedness communities are generally aware of the potential national security implications of scientific articles, scientists, physicians, and journal editors are much less aware of this threat. Identifying such research, its risks and benefits, and developing strategies for its communication are essential.

Following the September 11, 2001, terrorist attacks, anthrax mailings, and publication of several scientific articles raising security concerns, a group of journal editors and authors met in 2003 to discuss these issues (JAMA was represented in this group).² The group agreed on a statement encouraging (1) investigators to communicate results of research in ways that maximize public benefits and minimize risks of misuse and (2) editors to deal responsibly and effectively with the safety and security issues raised by scientific manuscripts and increase their capacity to identify such issues as they arise.²

Subsequently, the US National Research Council,³ the US National Science Advisory Board for Biosecurity (NSABB),⁴ and the World Health Organization⁵ developed reports and recommendations for identifying and managing DUR. The NSABB, a federal advisory committee composed of scientific, public health, security, and other experts was chartered in 2004 to provide advice, guidance, and leadership regarding biosecurity oversight of DUR. The NSABB defined a subset of DUR, dual use research of concern (DURC), as research needing special scrutiny: “research that, based on current understanding, can be reasonably anticipated to provide knowledge, products, or technologies that could be directly misapplied by others to pose a threat to public health and safety, agricultural crops and other plants, animals, the environment or materiel.”⁴ The NSABB developed detailed, practical guidance for communicating and publishing DURC results and provided examples of research categories warranting careful consideration for DURC: research findings on rendering vaccines ineffective, conferring resistance to antibiotics needed for biodefense, increasing microbial virulence, transmissibility and host range, enabling the weaponization of a pathogen or toxin, or decreasing the ability to detect pathogens.⁴ The NSABB makes recommendations to the US government but has no independent or regulatory authority; however, the NSABB's advice has considerable domestic and international influence.⁶

The unimpeded communication of valid and important research results is essential. The NSABB does not recommend rejection or censorship of manuscripts reporting DURC, but after due consideration of their potential effect on public health and security, to publish them, if at all possible, with any needed editorial modifications and accompanying contextual material (eg, an editorial or news release). This will help to ensure that the scientific and medical communities and the public are properly informed of the reason for the research and the meaning and significance of the findings. There may be an occasional case for which publication is contraindicated for security reasons. This is a decision made by the editor, not the government, unless the research is classified. Publishing a nonclassified article that might compromise national security would not violate any regulation or result in government-imposed sanctions.

The de novo chemical synthesis of poliovirus and reconstruction of 1918 pandemic influenza virus are examples of published DURC articles. The 2001 article⁷ describing increasing the virulence of mousepox illustrates the need for careful scrutiny and consideration of DURC manuscripts. The researchers used recombinant techniques to insert the gene for interleukin 4 into the mousepox virus and hoped the altered virus would induce infertility in mice (major agricultural pests in Australia) and serve as an infectious contraceptive for pest control. The investigators were surprised to find that the altered virus killed mice naturally resistant to and mice vaccinated against ordinary mousepox.⁸ Because mousepox is a member of the same family of viruses as smallpox, the unexpected results demonstrated how variola virus could be made more virulent through genetic manipulation, and the vaccine rendered ineffective. According to the senior author, peer reviewers did not identify any security concerns, and no editorial accompanied the article at the time of publication.⁸ Only when another journal published an article on the security implications of the mousepox article was worldwide publicity generated. Should that experiment have been performed? Should the results have been published? Should an editorial or news release have accompanied it? These are the types of questions that authors, editors, and reviewers should consider prior to publication.

There is no public information that terrorists or rogue states have thus far made use of the mousepox article or other published DURC articles to cause harm. This does not mean, however, that such articles have not been or will not be used to do so. Well-organized, valid information with the imprimatur of respected peer-reviewed journals could be especially valued by a malevolent actor over any information that might be available on the Internet.

A number of journal editors presented data on their experiences at a 2008 NSABB-hosted international conference.⁶ The Nature and American Society of Microbiology families of journals, Science, and the Proceedings of the National Academy of Sciences (PNAS) reported receiving few DURC manuscripts relative to the total number reviewed, and no manuscripts that needed to be rejected on the basis of security concerns from 2003 to 2008. Of the 74 000 biology submissions Nature journals received from 2005 to 2008, only 28 were considered DUR and none was determined to be DURC.⁶

The editor of Biosecurity and Bioterrorism reported rejecting DURC manuscripts when concerns could not be resolved. Three DURC manuscripts were received from 2002 to 2008.⁶ One, about aerosol delivery systems for bioweapons, was modified and published. A second described how powdered substances could be mixed with smallpox virus to confound the usual tests for detection. After the author refused to modify the manuscript to the satisfaction of the editor, the manuscript was rejected but subsequently published elsewhere. A third manuscript, modeling airborne anthrax attacks, discussed ideal weather conditions and how to release anthrax in buildings. The concerns were explained to the author and the manuscript was rejected.⁶

Scientific and medical journals that have not already done so should adopt policies to address the review, editorial evaluation, and decisions to publish DURC manuscripts. When manuscripts are submitted, investigators should identify potential security and public health implications of their manuscripts. With such policies in place, editors would be alerted to potential DURC, and authors would understand the possibility of delay, the need to revise, or rejection if DURC were identified and not properly dealt with.

In 2004, the World Association of Medical Editors released a policy statement entitled “Geopolitical Intrusion on Editorial Decisions.”⁹ The policy states that editors should consider whether publication of specific manuscripts “might cause harm to readers or to the public interest,” and that “decisions to edit and publish should not be determined by the policies of governments or other agencies outside of the journal itself.” More than 200 international medical journals have signed on to this policy.⁹ However, this policy does not specifically mention DURC. The American Society for Microbiology journals, Nature journals, PNAS, Science, and JAMA all have guidance for authors about manuscripts that may have security concerns. In a recent survey of 155 science journals, only 7.7% reported having DUR policies; 5.8% reported having had experience reviewing DUR manuscripts in the last 5 years.¹⁰

The current paucity of journal policies on DURC may be due to editors' lack of awareness, perceived irrelevance to what they publish, or concern about any interference in their decision making; perhaps they understand the need, but lack the time to deal with it.¹⁰ It may be time to convene another meeting of editors and researchers to address these concerns. At this meeting, existing policies of journals and associations of journal editors could be reviewed. Journals without such policies could adopt those best suited to them. In this way, the integrity and openness of scientific publishing would be maintained while potential security threats from the misuse of scientific articles may be prevented.