The H5N1 strain of influenza - often referred to as bird flu - was first known to have jumped from chickens to humans in 1997. Since 2004 it has ripped through Asian poultry farms, and had a 70% mortality rate in the first 70 people it is known to have infected. Health authorities fear this strain, or its descendents, could cause a lethal new flu pandemic with the potential to kill billions.
Flu has been a regular scourge of humanity for thousands of years. The flu viruses are a large family, each possessing a mere 10 genes encoded in RNA. All of the 16 known groups originate in water birds, especially ducks and gulls. The virus is well adapted to their immune systems, and does not usually make them very sick. This leaves the animals free to move around and spread the virus - just what it needs to persist.
But every now and then a bird flu virus jumps to an animal whose immune system it is not adapted to. In chickens, a forest bird originally and not a natural host, it causes a moderate disease but can readily mutate to a more severe strain. Just such a strain of H5N1 flu, named after its surface proteins, has been rampaging through large chicken farms in east Asia.
That is of concern because, in 1997, scientists found for the first time that H5 flu is capable of infecting humans. It was found in 18 people, six of whom died. All the poultry in Hong Kong were destroyed to stop the threat. But it continued to circulate, especially in China.
There were further human cases in China in 2003. Then in early 2004 Vietnam reported widespread poultry outbreaks and some human cases. After initial denials, Cambodia and Thailand admitted they had outbreaks too, followed by China. That was immediately after China had denied a New Scientist report that scientists strongly suspected Chinese outbreaks.
A mass poultry cull stopped the outbreaks by March 2004, by which time 23 people had died. But the virus persisted, most probably in ducks. Attempts to blame its persistence and spread in the region on wild birds are widely discounted by scientists. The outbreaks started again in summer 2004, and by mid-April 2005 had caused a total of 51 human deaths, all in Thailand and Vietnam.
The two or three flu virus families that have made the jump to humans mostly cause mild disease, because they have adapted to our immune systems. A yearly winter flu epidemic afflicts most of the world. But it is not totally benign. About 700,000 people around the world die of it each year, mainly the very old, very young and the infirm.
Common flu vaccines are increasing in popularity, although flu evolves so fast that we need new shots every year. In 2004 an unexpected shortage of vaccine in the US underscored the fragility of the vaccine supply, which is produced by very few manufacturers. As New Scientist predicted, it took strenuous efforts to limit available supplies to those most at risk of serious illness, preventing excess deaths.
But flu is at its most deadly when it first makes the jump to people, having had no opportunity to adapt itself to our immune systems. H5N1 has continued to infect humans as the outbreak in poultry has raged, with an apparently high fatality rate. It has so far has been hard to contract, and has not spread readily between people. If this viral strain should acquire that ability, it could become a lethal pandemic - the name for an epidemic that spreads worldwide.
That is what happened in 1918, when a virulent flu strain appeared in humans and killed 50 million people within a few months.
There have also been two less catastrophic pandemics. The so-called "Asian" flu of 1957 caused between one and four million deaths, while 1968‘s "Hong Kong" flu - with about half the estimated deadliness of the Asian flu - caused one to two million deaths. Both of these were human flu viruses which had recombined with bird flu viruses, rendering them unrecognisable to the human immune system. The 1957 strain was nearly released by accident in 2005.
Virologists generally agree that we are due for another pandemic. So they are very worried about H5N1, because - like the 1918 event - it seems to be evolving to become more deadly to mammals. This is largely in China and, possibly, as New Scientist revealed, in vaccinated chickens.
It could evolve into a potential pandemic that way, or by recombining with human flu, especially as most people in the Far East are not vaccinated against ordinary flu strains.
Fortunately we can make vaccines for the H5N1 strain, although our ability to get them tested and manufactured in time for a pandemic is in doubt. Once an effective vaccine is produced, yet another hurdle would be administering it swiftly. If either aspect of that process should fail, the only backup would be antiviral drugs. A few new ones are on the horizon, but existing drugs are in short supply.
If the flu virus changes genetically, it may become less deadly. However, there is no reason to think this will happen, and a highly contagious virus with a 70% death rate is a terrifying prospect, particularly given the speed of modern international travel. There is also a chance that it could evolve into a completely new disease, which we could fail to spot before it spreads.
Even if H5 does not trigger the next pandemic, its cousins H7 and H9 could. H7 is present in the same region and also infected large numbers of Dutch people in an outbreak in 2003. Although it caused few symptoms, and only one death, fears remain that such a poultry virus could cross-breed with a human flu, making it even more dangerous.
Some scientists are not willing to wait and see. They are trying to breed contagiousness into H5N1 to see if it is likely to happen. Others are breeding replicas of the 1918 virus - from samples recovered from victims - to see just what made it so deadly. But some feel that those experiments, because of the potential for escape from the lab, put us at as much risk as the natural evolution of the virus.
Debora Mackenzie, 6 May 2005
