Health - Doomsday flu decision time: The story so far

Last year, two flu research groups created what could be extremely dangerous viruses through their research into bird flu. Both studies will be censored when they are published, and all similar work has been put on hold – unprecedented actions in biomedical research. Ahead of a World Health Organization meeting to plot a way ahead, New Scientist explains how we have arrived in these uncharted waters.

What is the big deal with bird flu?
A highly dangerous flu called H5N1 evolved on chicken farms in China in the mid-1990s. By 2006 it had spread across Eurasia and as far as the UK and Nigeria. It is endemic and evolving in poultry in China, Vietnam, Indonesia, Bangladesh and Egypt, and it has killed 335 of the 584 people known to have caught it.

Those numbers are relatively low because for some reason H5N1 is very hard for humans to catch. The big question is whether it can mutate to spread easily among us while remaining as deadly.

So can it?
Several labs have tried to answer this all-important question by giving H5N1 various genetic changes they thought might make it spread easily among mammals. None of them did. Then Ron Fouchier and colleagues at the Erasmus Medical Centre in Rotterdam, the Netherlands, passed a mutant H5N1 virus repeatedly among ferrets, the best test animal for human flu. It picked up more mutations, which let it spread through air like ordinary flu, while staying just as lethal. We have an answer to the question, then.

Meanwhile Yoshihiro Kawaoka and colleagues at the University of Wisconsin-Madison put mammal-friendly mutations into the HA surface protein of H5N1 and tacked that onto a seasonal human flu in place of its H1 protein. That also spread easily through ferrets, though it barely made them ill.

And that's when the fuss started?
Alarm bells rang when the two experiments were submitted to journals. The National Science Advisory Board for Biosecurity (NSABB), a top US biosecurity committee, asked that the full details not be published, for fear bioterrorists would recreate the Dutch virus. They also worried that labs repeating either experiment might accidentally allow such a virus to escape.

Then the 39 top civilian flu labs in the world declared a halt to research aimed at making bird flu more transmissible, pending talks this month at the World Health Organization in Geneva, Switzerland. There, researchers and biosafety experts hope to agree how much of such work can be published, and whether withheld details should be shared with certain experts, for example so they can look out for potentially dangerous mutations in labs or in the wild.

Surely that last is a no-brainer?
You'd think so, but Paul Keim of Northern Arizona University in Flagstaff, the chair of the NSABB, disagrees. Keim, an expert on bacterial evolution who led investigations into the US anthrax attacks of 2001, said in an interview that there is little point looking for the virus that became transmissible in Dutch ferrets because it would spread faster than we could make vaccine for it.

However, Fouchier points out that if such a virus turned up in chickens, we could kill the chickens. Ab Osterhaus, the head of Fouchier's research group, says that even if the mutations created in Rotterdam aren't published, health agencies should still be able to watch for them in natural H5N1. One option may be to publish a longer list of H5N1 mutations that are of concern, without specifying which worked in Rotterdam.

Is there any chance a terrorist could make this virus and release it?
It seems unlikely, as it would threaten the terrorist's own people. Osterhaus says bioterrorism using the virus is unlikely because only a few dozen labs could make it. But no one can guarantee that deluded people won't try, which makes it risky to publish the recipe for such an extremely dangerous virus.

Surely there's little threat from the Wisconsin research? That virus didn't even kill the ferrets.
But the virus did spread readily among ferrets, and therefore might do so among humans too. If it got loose it would introduce the H5 surface protein to human flu viruses, which regularly swap genes. That might make another human flu virus much more virulent – and as an H5 virus has never circulated in people, no one would have immunity to it.

In fact, some virologists argue that we've never seen a natural H5 flu in mammals, so it might not be able to evolve in nature. If we help it past that barrier artificially, then release it, we could be making a problem for ourselves that would not otherwise have happened. Others argue that the research shows H5N1 can evolve this way naturally, so we need to try harder to eliminate the wild virus.