Understanding the Dutch export licence requirement for publishing life science research
During the Meeting of Experts of states parties to the Biological and Toxin Weapons Convention (BTWC) last August, the Netherlands organised or co-hosted three side events relating to safeguarding the life sciences. A significant incident, in which the Dutch virologist Ron Fouchier and his team were required to obtain an export licence to publish their research on how they had mutated H5N1 into an aerosol-transmissible avian influenza virus variant, undeniably informed the need to clarify national policies and approaches to biorisk management. A month earlier the Appellate Court had annulled the ruling by a lower court in support of the government position on procedural grounds. Does this annulment validate the Dutch government’s position or does it imply that the whole debate about the publication of so-called dual-use research in the life sciences is back to square one? Moreover, in the meantime the debate had evolved from a terrorist proliferation risk to one of health security in which the ethics and utility of this type of gain-of-function research stand central. In other words, do biosafety worries warrant biosecurity policy measures, such as the imposition of non-proliferation export controls?
Some background to the Netherlands decision
In September 2011 the European Scientific Working Group on Influenza (ESWI) held its fourth conference in Malta. Europe and the world were then confronting an outbreak of avian influenza caused by the H5N1 virus. Its rapid spread among birds over long distances caused governments worldwide to order drastic measures in efforts to stem the epidemic. Over 500 humans (representing some 60% of all people who had contracted the disease) had already died, but all deaths thus far had resulted from direct interaction with fowl and not from human-to-human transmission.
In Malta Ron Fouchier announced that he and his team at the Erasmus Medical Center in Rotterdam had succeeded in transforming H5N1 into a viable aerosol virus. According one conference report he applied rather colourful language: his team ‘mutated the hell out of H5N1’. The discovery that it required as few as three single mutations to gain the ability to latch onto cells in the nasal and tracheal passageways, he described as ‘very bad news’. Transmission among ferrets, a mammal that offers the best laboratory model to study influenza in humans, still did not occur easily when, in Fouchier’s recorded words, ‘someone finally convinced me to do something really, really stupid’. They provoked two further mutations by transferring mutated viruses from the nose of one sick ferret to that of a healthy one, in the process creating the viable aerosol virus.
Initial articles on the gain-of-function research did not suggest any link with bioterrorism, but sometimes carried dramatic titles evoking cataclysmic consequences reminiscent of the 1918 Spanish flu epidemic that killed tens of millions worldwide. However, Fouchier’s dramatic speech caught the attention of counter-terrorism officials on both sides of the Atlantic. When he offered his research results for publication in Science, the US National Science Advisory Board for Biosecurity (NSABB) intervened and eventually accepted publication provided some methodological details were removed from the text. (A parallel paper on H5N1 submitted to Nature by a team led by US scientist Yoshihiro Kawaoka fared a similar fate.) NSABB considered the biosecurity risks outweighed any scientific merit in this type of research: rogue laboratory researchers or terrorists might wish to unleash the deadly virus on the human race to devastating effect.
NSABB’s intervention caused controversy with one side calling it censorship and the other side up in arms that publication had been authorised at all. In the United States, self-publication on the internet was not legally restricted at the time. In the Netherlands too the affair had caught the attention of authorities, many of whom wanted to block publication outright, but lacked the appropriate legal tools to do so. The only way to appraise the risks for malfeasance posed by information in the research manuscript was to implement non-proliferation export controls, which according to European Union regulations must be enforced for applied (but not fundamental) research in the life sciences.
The Netherlands decision caused shockwaves among European life scientists. After initial defiance, Fouchier and his team eventually applied for and received an export licence. Otherwise he might have faced up to six years imprisonment and $102,000 in fines. The Erasmus Medical Center subsequently took the government to court to have the principle of export licences for scientific research overturned. On 20 September 2013 the District Court of North Holland ruled in favour of the government, but on 15 July 2015 the Appellate Court in Amsterdam annulled the ruling, saying that the case was without merit in view of the application and granting of an export licence. Put differently, the lower court should have never taken up the case. As presented during the BTWC Meeting of Experts in August 2015, the Dutch Government believes that the juridical process vindicated its approach to research with potential dual-use implications.
Assessing the implications
In August I spoke with Koos van der Bruggen about the impact of the Dutch export licence requirement on the life sciences. Koos has had a privileged position to observe the unfolding of the debate. With a PhD in political science, he worked for more than 12 years at the Rathenau Institute, the Dutch organisation for parliamentary technology assessment. He was involved in research and debates on social and ethical aspects of technological developments, including biomedical technology, genetics and military technology. Later, while working at the Royal Netherlands Academy of Arts and Sciences he participated in the development of a code of conduct for scientists regarding biosecurity issues. At the Delft University of Technology he conducted research on biosecurity and dual use. Today he is an independent researcher on peace and security matters and often solicited in the Netherlands and abroad on issues of ethics and technology relating to biological weapons.
Can you briefly summarise the various elements that came into play in the Fouchier case?
People say that the case really started with his presentation at the Malta conference in 2011. In fact one can argue that it began 4–5 years earlier when Fouchier requested and obtained NIH (National Institutes of Health) funding to investigate whether mutation of the H5N1 virus was possible in a way so that it would become aerosol transmissible. At that moment the laboratory already had to comply with many rules. It had its status upgraded from BSL 3 to BSL 3+ to enable this research. Fouchier not only had to deal with rules in the Netherlands, but also with those in the USA because of the NIH funding. So, all biosecurity and -safety elements had already been considered in that phase. He then started the research with his group. I still think that he and his researchers did the job as they thought it should be done. They were surprised when it appeared that they had succeeded in mutating the virus in such a way that it became aerosol transmissible.
How would you characterise fundamental research in the life sciences?
That is a difficult question to start with. I sometimes ask myself if fundamental research still exists in the life sciences. Of course, there is research into the essence of bacteria, stem cells, and in other fields. Nowadays, almost all science is directed at results, at getting the costs back. Already from the start researchers are thinking about applying the science.
The question if the H5N1 research was fundamental of course played a role in the debate on the Fouchier case in the Netherlands because fundamental research is exempt from the EU export license regulations. It was therefore not surprising that Fouchier claimed that his work was fundamental research, although at other moments and in other places he just said his research was directed at finding ways to prevent the bird flu or finding vaccines for treating the bird flu. So, that was not so coherent.
At the time, when he was getting his results, the question of fundamental research did not arise in any significant way. Would you agree that the determination that a particular scientific result may lead to a non-conventional weapon affects the definition of fundamental research?
I do not think that life scientists spend one second making such a link. They even hardly do so when engaging in applied research. Weapons of mass destruction are not on their mind. It is difficult to imagine that they would link their work to weapon development. Scientists should be more aware of it, but it would be an overreaction if it were to influence their daily work.
Research is governed by many protocols on safety and security. There is discussion about the research purpose, the adequacy of facilities, and so on. In the case of H5N1 project funding came from the United States. I imagine that the project was governed by regulations from the Netherlands, the European Union, the United States, perhaps even certain regulations set up by the WHO. With all these elements in place, what is their value if all can be blocked by a judicial decision?
I do not know if everything is blocked. I agree when you say that Fouchier had to comply with all kinds of regulations from the USA, Netherlands, EU, WHO, etc. As far as I can see these rules were never the problem. Nobody told Fouchier that he had not complied with any particular rule. So, he could do the research.
The discussion started when he wanted to publish the results. For scientists publications are nowadays essential. I can imagine that he felt uneasy about this outcome. It surprised him because he had complied with all relevant rules and regulations.
A big difference has emerged between laboratory biosecurity and biosafety and what we might call ‘informational biosecurity‘. This is indeed about publishing, communications at conferences, and so on. Other rules than laboratory safety and security measures then come into play. To me, that is one of the essential differences, which became visible in this debate. People did not realise its significance before.
Elsewhere you have put forward the interesting questions: how is dual-use research assessed and who determines if research is dual use? The key question that comes to my mind is: what defines ‘dual use’ in research? I can see two approaches. Either it is an intrinsic characteristic, in other words, it is inherent to the project. In that case, everything becomes problematic one way or another from the perspective of potential armament programmes. Or, it is an extrinsic characteristic, in other words, a potential that exists, but which will only be realised under certain circumstances. In the concrete case of the H5N1 research project that determination happened when the results were already available. One proposal put forward by you and some other experts holds that such considerations should be taken into account at the funding stage. So, even if one shifts the whole to debate up front of the project process, how would one decide it is dual use?
I agree with your characterisation of dual use. If everything is dual use, then my pen is a weapon, because I can stick it in your eye. This is not the debate. I refer to the report by the Netherlands Royal Academy of Arts and Sciences, which is more in line with your second description that dual use is dependent on circumstances. We should not only look at the physical or biological characteristics of pathogens – then we would get into your first description of dual use – but also at political, social and other circumstances that can make a biological agent dual use. The problem, of course, is that one cannot always predict those circumstances. Doing this H5N1 research in the safe Erasmus laboratory in Rotterdam is not a problem, but doing it today in Damascus, Syria would be extremely problematic. Same research, different places. That is one point. The other point is that in good science, one does not know exactly what may happen. Researchers work on the basis of hypotheses, and these may be falsified. Fouchier and his collaborators were surprised by their outcome. We advised to pay attention to these possible dual-use dimensions from the beginning in the funding phase, but also to bear them during the whole project implementation phase in mind. The initial assessment may change for several reasons, for instance, if results differ from expectations, or if the social or political circumstances shift. I realise that makes ‘dual use’ not very tangible.
Basically you are pleading that scientists should have enhanced social awareness.
They should be socially aware, but you cannot expect that they know all these things. This is why we advised to have a permanent commission that can help scientists especially in those areas where they are not experts themselves – the social and political aspects, the security dimensions, and so on. In October 2014 the Dutch government decided not to adopt that advice. They think that the Export Control Office of the Ministry of Foreign Affairs and the specialised Biosecurity Office have enough expertise and experience to do so. The Academy regrets this decision, Of course there is expertise in these offices, but (1) dual use is more than export control and (2) both offices are part of government, so not independent.
That special commission, would it have functioned for the entire duration of the project or only at certain stages?
The idea was that scientists could go to that commission whenever they thought it might be necessary. In our view, the primary responsibility still rests with the scientist. However, he should know that if he does stupid things along the road, government instances may tell him that he cannot publish his results or must terminate the research project.
The question of the H5N1 study hit the headlines in 2011, but the project began earlier. Already in 2007 the Netherlands had adopted a code of conduct for biosecurity. Did it have any impact on the H5N1 research project?
The curious thing is that Ron Fouchier was on the advisory board for this code of conduct. He knew about it. I also know that he used it when starting this research project on H5N1. One of the points – and I think it relates to the definition of ‘dual-use research of concern’ as they say in the USA – is that it is not that tangible. He was convinced that he could carry out his research in a way compatible with the rules in the code of conduct. Many people have said ‘strange that he was in this code of conduct commission and still he did so‘. I think that gives a wrong impression of what a code of conduct is. It is intended to make people think about issues, make them aware of them, but it does not give concrete answers to concrete questions such as whether this or that kind of research is acceptable or not.
The code of conduct was revised in 2013 …
It was not really revised. We had a kind of evaluation and decided not to formally revise it. I think that if we had written it in 2013 some articles would have been different, but not in such a way that it would have required a revision.
At the various stages of the legal process did the judges take that code of conduct into consideration when writing their rulings, or did they go straight for the export control regulations?
I read the verdicts and I did not see any reference to the code of conduct.
The world-wide debate initiated by the H5N1 gain-of-function research has now shifted from the terrorism paradigm, what you have called the biosecurity angle, to the global health security, which one could call the biosafety paradigm. This is a totally different framework for considering the concerns of a couple of years ago. The recent ruling from the Netherlands Appellate Court still stands. Did the judges reach their conclusions in function of the policy concerns of the day?
This is difficult to say. The judgement of the Haarlem court, that is the first judgement of September 2013, was in support of government policy. Its arguments referred to Security Council Resolution 1540, non-proliferation considerations, and so on. The conclusion was that Dutch government policy was the right policy, and therefore that the government was right in demanding an export licence for this research article. The Appellate Court, as I see it, made a purely legal and procedural kind of argumentation. They said that the Erasmus Medical Center got the export licence, so they were not entitled to have an appeal because they already got what they wanted to have. This means that the judgement of the Haarlem court should not have been done in this way. The Haarlem court should already have said that there was no case.
In practical terms, the ruling by the higher court means that the judgement by the Haarlem court is also without substance. Maybe the plus point for Fouchier might be that the export licence requirement is now also off the table. Nevertheless, key is that the Appellate Court sidestepped the substantive part of the question and merely issued a procedural verdict.
Are you now implying that there was no determination whether the research was fundamental or applied?
… or whether the government was entitled to ask for an export licence or not. The Appellate Court in fact said that if there were a new case, one could try the legal process again in case of refusal of an export licence or even if one does not want to apply for an export licence.
Are you now saying that given the ruling that the Haarlem verdict is without substance, researchers involved in a similar project with similar outcomes could send off their findings for publication without requesting an export licence? It is hypothetical, but …
I do not think that the scientists and their lawyer came to that conclusion. The common opinion is that the same rules still apply as in 2011: asking for a licence is necessary. And if you do not want to do so, you can go to the court before and not after asking for a licence. But what might happen if a researcher decides to just publish and see what happens? I do not know if he would dare to do so.
Well yes, in terms of publication time lines, the process could take several years. Now, if we look at the content of the Haarlem ruling and that of the Amsterdam decision, is it now clear if an export licence would be required in the case of a publication in a Dutch or European Union journal? In his case, Fouchier went to the United States.
Formally, if you publish your findings in, say, a Dutch Journal of biology, for instance, with a Dutch owner and within Dutch borders, even perhaps within the EU, you should not have to ask for a licence. Practice, of course, is that it will be spread around the world maybe within minutes or perhaps a few days. As a consequence, an export licence may therefore be required. For many things the issue would be clearer, but for digital content it is very different. As far as I know, the judgements of both courts do not say anything about this. I do not know how the government thinks about this. But it should have given it some thought, because that would be the simplest solution for all researchers.
The logical follow-on question to your answer is: what about other forms of communicating results? Researchers go to conferences, to seminars. There is the internet. One cannot go to a conference requesting an export licence for what one is going to say. What about conversations at the bar?
I know that before the time of the judgement Ron Fouchier would no longer talk about the results, at least not about the details that were in question. But he had already done so in Malta where he announced the first results.
In fact, the same about what I have already said about publishing in journals also applies to conferences. It is very difficult for scientists to have a lock on their mouths for certain aspects of their research. On the other hand, it sounds a bit ridiculous to go, as you say, with an export licence to a conference in Malta or wherever, for sending e-mails, and all these kinds of things.
Since the 1960s and 1970s the Netherlands has widely debated the autonomy of science, the social responsibility of scientists and the relationship of scientific research to society. The debate was often expressed in different ways, for example, the refusal of grants from entities deemed unsuitable—NATO was one that cropped up often—or insistence on the right to scientific cooperation with political regimes considered to be adversarial during the Cold War. Yet, despite this openness universities also had to accept a Crown representative or representatives on their board to add certain control over the choices in scientific research. In the middle of the 1970s there was also a debate about the reorganisation of government-sponsored research. Are we seeing a reinforcement of a policy trend that started emerging in the 1970s, but is now driven by considerations of biosecurity and terrorism?
I see great differences in university policies certainly since the 1990s, because universities have become more commercial, more international also. They started working together more with other universities across the world. I was a young scientist in the 1970s and 1980s, so have been certainly involved in these kinds of debates when I was at the Rathenau Institute, the Dutch institute for technology assessment. I remember indeed that the so-called critical scientists that we were in those days opposed certain kinds of research, for instance research that in one way or another contributed to nuclear weapons. Research done for NATO was also out of order. This critical sense has perhaps been lost today, but some other developments have occurred. The Rathenau Institute is an example. More attention has been paid to ethical aspects of science and technology. These aspects had to be considered explicitly in applications for projects in the life sciences or the natural sciences. Because of that, the discipline of Ethics has grown in the Netherlands. In many projects an ethicist has become part of the research team.
So what you are saying essentially is that there has been a parallel evolution of societal responsibility of the scientific community. This does not have to do explicitly with government control.
No. It came from the scientific community itself. But these developments were also supported by government. But their support did not take the form of government interference with science policy. This is why the H5N1 debate surprised so many scientists. The first reaction of the then president of the Royal Netherlands Academy of Arts and Sciences, Robert Dijkgraaf, was indeed very heavy. He and the other members of the Academy later regretted the tone of his reaction. But they did so because the measure came as an absolute surprise. It was seen as a a kind of censorship.
Well, this brings me to my last question. Now we are 2015 and there has been development in the Fouchier dossier from a legal point of view, but there has also been a whole evolution in the presentation of the problem. Where does it leaves us with regard to the publication of research, the ability to verify both research methodology and research results to be able to reproduce them? There is a lot of research fraud going on; the number of journal articles being retracted is increasing. The freedom to develop research projects … where has the H5N1 dossier left us?
I do not know whether I can answer this question. I think in general that the impact of the H5N1 case may not have been as great as what people believed in 2011–2012. I think most of the research can still go on, although I also hear of researchers who, particularly in the United States, do not even start projects with certain pathogens, such as anthrax bacteria, because they know there are so many hurdles to overcome even before they can start. This is not a good development. Of course, there is the impact of the US moratorium on funding of some kinds of gain-of-function research. Given that the US is the main funder of many kinds of this type of research, this also has a broader repercussions on all research as many projects are now suspended. That really has an impact. Many researchers ask why specifically this kind of research has been suspended. Why this field? In my view this is a direct consequence of the 2011–2012 debates. But it does not always seem logical. Fouchier has said several times that he conducted research that was much more dangerous in his eyes than that on the H5N1 virus. So, that is also an issue.
It was a specific set of circumstances that brought about the reaction to the H5N1 research.
Yes, it was a coincidence of factors.
Would it be a fair assessment to say that al least in the Netherlands publication was the problem and not the research as such?
I think so, and probably also in the United States. It started with the debate on publication. As far as I know, nobody posed questions about the research itself during the early days of the debate. That came later, as you have said earlier with the shift from biosecurity to biosafety and the questioning of the value of this kind of research. Then some virologists – Simon Wain Hobson, Marc Lipsitch and others – have claimed that this is not good research. But then we are in another field of discussion, not in the security field. As a political scientist I am not in a position to say anything useful in that discussion.
Thank you very much for the interview.