One type of the influenza A virus occurs naturally in wild waterfowl. It has at least 16 sub-types, which differ in their haemagglutinin (HA) and neuraminidase (NA) surface proteins. When exploiting new host types – either avian, e.g. chicken, or mammalian, e.g. human – the virus undergoes rapid change.

This influenza A virus usually stays in ‘harmless’ equilibrium inside the natural host with minimal evolution at amino acid level, which is surprising for a RNA virus highly prone to variation. Virologists think benign equilibrium may have been overtaken by dynamic change.

Prior to 1997, there was no evidence for H5 influenza A virus infecting humans with fatal consequences. Precursor of the H5N1 virus that infected humans in Hong Kong in 1997 was first identified in geese in Guangdong in southern China one year before. Before hitting Hong Kong’s poultry markets and killing 6 out of 18 persons infected, the ‘goose H5N1’ acquired internal gene segments from H9N2 found in quail and a neuraminidase gene segment from an H6N1 duck virus isolated from the common teal, Anas crecca.

This 1997 H5N1 virus strain was eradicated by culling the entire Hong Kong poultry flock. The strain was not seen again but other re-assortments continued to emerge from virus reservoirs in China’s duck and goose populations.

One such strain hit Hong Kong in late 2002, killing wild and domestic waterfowl and infecting several people with one fatality. It subsequently proved to be precursor for the ‘Z’ genotype which spread like wildfire across Thailand, Vietnam, Indonesia and other south-east Asian nations in early 2004. South Korea and Japan were also hit, but with a different strain called the ‘V’ genotype.


Further evolution within the ‘Z’ genotype showed that viruses in Vietnam and Thailand had formed distinct clusters distinguishable from those in Indonesia. All showed a capacity to cause neurotropic disease and high death rates in wild waterfowl, transmission to and mortality in cats and ability to infect and kill humans. Genetic changes allowing these new capabilities initially came about by re-assortment. Later changes were through mutations in the RNA polymerase (PB2) gene, insertions in the HA gene and deletions in the NA and non structural (NS) genes.

H5N1 had killed waterfowl in Hong Kong in late 2002 and the H5N1 viruses isolated in Vietnam, Thailand and Indonesia thereafter were highly pathogenic to domestic ducks as well as chickens. Later on, research showed the selection within ducks of a subdominant population of H5N1 viruses that appeared after the immune response had cleared the dominant virus. The subdominant population appeared consistently non-pathogenic to ducks as if this was the natural situation for influenza A virus in the duck.

H5N1’s loss of pathogenicity in ducks while maintaining high virulence in chickens and humans is the root cause of past and present eradication problems in Vietnam, with its huge and mobile duck population. Disease symptoms in ducks were initially used in Vietnam to identify H5N1 but this development now makes it impossible.

In a thorough review, Dr Robert Webster* of St Jude Children’s Research Hospital in Memphis, USA, says the duck has become the ‘Trojan Horse’ for highly pathogenic H5N1 influenza in Asia.

* Webster, R.G. et al., 2006. H5N1 outbreaks and enzootic influenza. Emerging Infectious Diseases, Vol. 12, No 1, pages 3-8.