U.S. committee will decide if international scientists allowed to publish flu research

The row over whether scientific journals can publish details of mutant strains of the H5N1 bird flu virus that can spread to other animals is about to come to a head in Washington

One Monday morning in September last year, Ron Fouchier, a virologist at the Erasmus medical centre in Rotterdam, stood at the Intercontinental hotel in Malta and told an audience of scientists how he created one of the world's most dangerous viruses.

In a secure laboratory built to contain harmful pathogens, Fouchier took the H5N1 bird flu virus and mutated it, through some worryingly simple steps, into an airborne strain that spread swiftly among ferrets in neighbouring cages. At first glance, the work might seem bad news for ferrets and little more. But the animals are the best mimic we have for how the virus could infect people, for example, through coughs and sneezes.

The work went largely unnoticed until December, when the US government's biosecurity watchdog raised the alarm. Though bird flu, as the name suggests, is mostly an avian disease, it has killed more than half the people known to be infected. The only reason it has not become a global killer is that it does not spread easily from person to person. The National Science Advisory Board for Biosecurity (NSABB) declared a paper Fouchier had sent them – and the US journal Science – too dangerous to publish. The board called for key sections of the report to be deleted, to prevent the recipe for the virus falling into the hands of bioterrorists.

Fouchier was not alone in having his work black-marked. The NSABB urged similar restrictions on another paper sent to the British journal Nature. In that, Yoshihiro Kawaoka, a virology professor at the University of Wisconsin-Madison, merged bird flu with the highly transmissible swine flu virus, to make a hybrid strain that also spread among ferrets through airborne droplets. The swine flu virus, which originated in pigs, went global in humans, but while it killed more than 18,000 in the 2009 pandemic, was not as lethal as the World Health Organisation (WHO) feared.

The advisory board's reaction has sparked a rare crisis in science. The US government backed the NSABB, but many researchers say the work must be published in full, arguing public health will benefit. A group convened by the WHO recommended full disclosure, but ordered an urgent review of the security and safety of labs where such viruses are stored. In the meantime, work on mutant bird flu has halted under a voluntary moratorium.

The deadlock could soon be broken. On Thursday, the 23-strong NSABB gathers for a confidential two-day meeting in Washington to review updated versions of both papers, which in Fouchier's case includes fresh evidence that his virus is not as dangerous as some accounts first claimed. Fouchier and Kawaoka will be brought in to answer queries directly. Whatever the panel's final recommendations, they will redefine how risky, dual-use research is handled in future.

There are solid scientific grounds for tinkering with the virus. Fouchier and Kawaoka's experiments address a crucial question: could bird flu evolve naturally into a form that spreads easily among people? Their work suggests it can. In Fouchier's study it took only five mutations – all of which have been seen in the wild – for bird flu to become airborne and spread through ferrets. Kawaoka showed bird flu could achieve the same end by picking up other genes when it mixed with swine flu, an event that could happen naturally in pigs.

The search for a resolution to the impasse has been clouded by several factors that created a perfect storm over the issue. First, fears that a bird flu pandemic in humans might be extraordinarily lethal, perhaps more so than the Spanish 1918 flu virus that killed only 2%-3% of those infected, but still claimed tens of millions of lives. Officially, 59% of the 600 or so people known to have caught bird flu have died. But that figure is likely to be misleading, as mild cases might go unrecorded. Just how lethal a bird flu pandemic would be in humans is unknown. Second, the threat of bioterrorism. And third, comments in the press.

'One of the most dangerous viruses you can make'

In November, Fouchier told Science his strain of flu was "probably one of the most dangerous viruses you can make". The head of the NSABB, Paul Keim, who has worked extensively on anthrax, told the journal: "I can't think of another pathogenic organism that is as scary as this one. I don't think anthrax is scary at all compared to this."

Secrecy over the research added to widespread confusion in the media over Fouchier's virus in particular. While Kawaoka's strain was not lethal to ferrets, several reports claimed Fouchier's was. He later clarified that the airborne virus only killed ferrets when it was squeezed in large doses directly into their tracheas. Some animals that caught the virus by inhaling it from others got sick, but none of them died. "Because of the confidentiality of reviewing the work, these details were not entirely clear and there were a lot of misconceptions out there. The lay public was really confused," said Anthony Fauci, director of the US funding agency the National Institute of Allergy and Infectious Diseases, who called the NSABB meeting on Thursday.

As a test case, the mutant flu papers have exposed what some consider serious shortcomings in the mechanisms that governments, funding agencies, scientists and journals have in place to handle potentially dangerous discoveries. The NSABB, for example, was oblivious to the work until October when the journals were ready to publish the papers. "It's a lot of work to get an advisory board fully informed about specific research. To do it quickly and within the scientific publishing schedule is next to impossible," said Keim. "The US government came close to killing us with the workload last fall. One committee member estimated that he devoted 200 hours to the problem in October and November."

The NSABB's call for redactions revealed other systemic problems. They wanted journals to publish the research with sensitive information removed. Only an approved list of scientists could then apply to read the sensitive information.

"It's a nonsense. Who chooses the 200 or 400 scientists around the world who get access? Who polices whether they immediately give it to their colleagues? It's unworkable, unpoliceable and crazy to even consider. Once it's out there, it's out there," said Jeremy Farrar, director of the Oxford University clinical research unit in Ho Chi Minh City, Vietnam.

When it comes to bird flu in humans, Farrar's clinic is on the front line. Staff there treat patients who contract the virus and run surveillance for unusual strains in people and poultry, pigs and other animals. This month, two men turned up in Ho Chi Minh City with bird flu infections that looked different to all the previous cases Farrar has witnessed in the past seven or eight years. The men, in their 20s, arrived from the countryside desperately sick. Tests confirmed they had H5N1 bird flu and X-rays of their lungs showed white, a clear sign of the damage the virus can cause. The outlook was bad for both, but within days, the men had recovered. The unusual cases raised questions for Farrar.

"What you are always looking for is a change in the way the disease presents itself and how the body responds. It makes you question whether the disease is adapting to humans. It's pure speculation, but we could be seeing the process of adaptation going on in front of our eyes," Farrar said.

The genetic mutations that made Fouchier and Kawaoka's viruses so transmissible in ferrets would not necessarily be the same ones that help bird flu jump into humans. But if the details were published, Farrar said he could at least screen for them and learn whether the mutations appear only singularly in the wild, or start appearing in clusters of twos, threes and fours. More importantly, knowing how the mutations transform the virus would help scientists spot other mutations that could make bird flu adapt to humans. "All of this surveillance is not much value if the experimental work, which is mostly done in western labs, is not made available to the countries where it's most needed. We can't be blinkered into thinking these are the only mutations that matter, but the information could be important for us," Farrar said.

One difficulty facing bird flu research is that surveillance picks up only a tiny fraction of the millions of viruses at large in the world. The chances of spotting a pandemic strain early enough to contain it are slim. For that reason, many scientists say the best we can do is prepare for the disease when it strikes.

"Forget all the nonsense about bioterrorism. These papers tell us that nature is where we should focus our concern," said John Oxford, professor of virology at Barts and the London School of Medicine and Dentistry. "The mutant strains that Fouchier and Kawaoka described are probably already out there. They must be out there. Fortunately for us, it is probably in some duck in Siberia, but were it to move close to hand by the vagaries and chances of nature, we'd be for it."

Governments should prepare

Oxford and Farrar argue that the mutant flu papers stress the need for governments to prepare for a dangerous, emergent flu strain. The Department of Health has 16m doses of the GlaxoSmithKline bird flu vaccine, Prepandrix, stockpiled with a shelf life of three to seven years. If bird flu jumped into humans, the frontline services, including doctors and nurses, would get priority for the jabs.

Ab Osterhaus, a leading authority on flu and head of Fouchier's department in the Netherlands, said the risk of a terrorist attack had to be weighed against the benefits of circulating the information so more scientists can work on combatting dangerous mutant flu strains. "Of course there is a theoretical possibility that these data will get into the hands of people who really want to do bioterror attacks, but if you look back in history, what kind of bioterror have we seen? We've had flu pandemics, Sars, HIV. The real terrorist is nature," he said.

After the NSABB raised the alarm over Fouchier and Kawaoka's papers, Fauci urged the WHO to take a look at the work. On 16 February, 22 people in the WHO's Geneva offices to thrash out a consensus. Fouchier, Kawaoka and Keim were there, with other flu researchers, journal editors, and representatives from Indonesia and Vietnam, where bird flu is endemic. The group decided there was too little time to create a global mechanism to disseminate sensitive details from redacted papers to a chosen list of experts. Instead, they backed full publication, but only after a public awareness programme to explain the importance of the research, and a review of the safety and security of mutant flu strains.

Though Keim was at the WHO meeting, he says its recommendations will not dominate the NSABB's own deliberations this week.

"The WHO meeting was good in that it was the first international meeting to weigh in on this subject. But it was a very limited group of international experts drawn from WHO influenza collaborating laboratories. I respect the opinion of the group, but I strongly believe that policy in this arena needs to be made by more than the flu research community," Keim said. "The cradle-to-grave management needs to involve outside and independent individuals. The institutions involved, including the National Institutes of Health (NIH), claim they were aware of this research and that they followed all the recommended procedures. If so, why are we in this situation?" he said.

Many scientists at least agree with the WHO's call for an urgent review of the safety and security surrounding mutant flu research. Laboratories where pathogens are handled are built to a biosafety level (BSL) ranked from one to four. Lethal agents, such as the Ebola virus and smallpox are contained in the highest-level laboratories (BSL4). Both mutant flu studies were performed in BSL3 labs, which are considered more than adequate.

Even so, some countries have already upgraded their security around mutant bird flu work. Last month, Canada ruled that all future research on strains of bird flu that are transmissible in mammals be confined to BSL4 laboratories. The prospect of a mutant flu virus escaping from a laboratory is not far-fetched. In 2009, scientists at the University of Pittsburgh wrote in the New England Journal of Medicine that the 1977 flu pandemic was probably released from a lab by accident.

"The institutes, the funding agencies, the scientists and the public need to be convinced there is absolute biosecurity so those viruses don't leak out. That's a major concern, and it should have been dealt with at the time the work was funded," said Farrar.

At the US National Institute of Allergy and Infectious Diseases, the division of the NIH that supported the mutant flu work, guidelines are now being drawn up to ensure formal discussions take place with scientists about the direction of their work and how to deal with potentially dangerous discoveries before the work is done. "We are putting together a much more formal policy of how we will address dual-use research of concern. They are not going to be unspoken assumptions any more, they're going to be official," Fauci said. "But there is a delicate balance. There cannot be any impediment to science that will ultimately be good to the general public."

Decision due on Friday

The NSABB's final decision, due on Friday at the earliest, will be considered by the US government before being passed to Science and Nature, whose editors have expressed a preference to publish the papers in full. Whatever the fate of the mutant flu papers, moves towards a mechanism of distributing sensitive scientific information in future are likely to get a spur.

"The important message is that what has happened here should never happen again. In the future, there will be work that is important to do, but it might be done in a classified setting, where the safety is assured and the information is shared only with those who need to know. This is not just about influenza. What is next agent coming down the line? Today we are in a whole new world of microbial genetics, and this is going to happen more and more," said Michael Osterholm, an NSABB member at the Centre for Infectious Disease Research and Policy at the University of Minnesota.

Fouchier fears the NSABB will pave the way to a precedent in censoring scientific publications and warns that an international agreement to allow selective distribution of sensitive information will be "laborious and time-consuming".

The Dutch government has already insisted Fouchier applies for export licences for his paper to be discussed at the WHO and today's NSABB meeting. It has threatened to block publication of his paper unilaterally through the same legislation. Any country signing up to a secretive distribution system might have to rewrite its own laws on the export of information.

Efforts to draw up a distribution mechanism at the NIAID have made little headway. "We've been trying to work it out, just in case the journals have to publish redacted versions, but we're still trying to figure that out. The problem is, how do you give access to unredacted information?" Fauci said.

Fouchier has similar concerns. "If the scientific community is expected to work via redacted papers and sharing of scientific information on a need-to-know basis in the future, then international governments better start to work towards systems to facilitate that. While this system is not in place, advice such as that from the NSABB last year simply cannot be followed," he said.

• This article was amended on 28 March to correct Michael Osterholm's name

guardian.co.uk © Guardian News and Media 2012