The virus that caused the 1957 "Asian flu" pandemic has been accidentally released by a lab in the US, and sent all over the world in test kits which scientists are now scrambling to destroy.
There are fears the virus could escape the labs, as the mistake was discovered after the virus escaped from a kit at a high-containment lab in Canada. Such an escape could spread worldwide, as demonstrated in Russia in the 1970s.
The flu testing kits were sent to some 3700 labs between October 2004 and February 2005 by the College of American Pathologists (CAP), a professional body which helps pathology laboratories improve their accuracy, by sending them unidentified samples of various germs to identify.
The CAP kits - prepared by private contractor Meridian Bioscience in Cincinnati, US - were to contain a particular strain of influenza A - the viral family that causes most flu worldwide. But instead of choosing a strain from the hundreds of recently circulating influenza A viruses, the firm chose the 1957 pandemic strain.
This is a problem because of the way pandemic flu strains edge each other out of circulation. The most lethal flu pandemic on record, in 1918, was caused by an influenza A of the H1 type, named for the haemagglutinin, a surface protein, it carries. After 1918, H1 flu evolved into an "ordinary" flu, and continued to circulate.
The 1957 pandemic started in China before spreading worldwide, killing an estimated two million or more people. It was triggered by the hybridisation of human H1 flu with flu viruses from birds which carried another surface protein, H2. It was more lethal than the then-circulating H1 strains because no human had ever encountered the H2 protein before, and so lacked any immunity to the new strain.
Immediately after 1957, all traces of H1 flu in humans disappeared, to be replaced by H2 strains. A similar process occurred again in 1968, when another hybrid virus emerged - again in China - carrying another haemagglutinin, H3. This caused the "Hong Kong flu" pandemic, which killed an estimated one million people worldwide.
But after 1968, H2 flu disappeared - so anyone born after this year will have no immunity to H2 flu and any escape of the virus in the test kits could be as lethal to them as the Asian flu of 1957.
A similar event happened in 1977, with the sudden reappearance of an H1 flu identical to one that had been isolated in 1950. It is believed that the virus escaped from a faulty batch of live flu vaccine prepared in Russia. But fortunately that strain had evolved into a much tamer creature than its 1918 predecessor. Unfortunately, the 1957 H2 virus is the most lethal variant of its kind.
A few of the CAP kits were sent to labs in Asia, the Middle East and South America, as well as Europe and North America. The kits' originators should have known what strain they contained, in order to evaluate the test results, though they claim they did not realise their mistake.
However, when Canada's National Microbiology Lab in Winnipeg identified the strain on 26 March - in a routine sample sent there from a Vancouver-based lab - it alerted the US Centers for Disease Control and the World Health Organization.
A major concern is that test kits are not usually handled at a high level of biological containment as it is generally assumed they do not carry unusually dangerous viruses. The Asian flu's most probable route of escape into the outside world would be if a lab worker were to unknowingly become infected by it.
But there has been no sign of the virus infecting humans yet, says Klaus Stöhr, chief flu scientist at the World Health Organization in Geneva.
"If this incident doesn't cause a major reassessment of the safety of flu research, a lab-sponsored pandemic may well be the only thing that induces sobriety," comments Ed Hammond of the Sunshine Project, a biosafety pressure group. (New Scientist, 4.13. 2005, Debora MacKenzie) http://www.newscientist.com/article/dn7261