Dept title


Current state of knowledge on Highly Pathogenic Avian Influenza

2.2 Spread of H5N1 HPAI in Asia and Beyond

Transborder spread of H5N1 HPAI was first reported in late 2003 and early 2004, yet H5N1 HPAI viruses were already present in Viet Nam (where the first H5N1 HPAI isolates were detected in 2001) and China (1996).  Even by 2002, numerous Asian-lineage H5N1 HPAI virus genotypes had already emerged (Sims et al, 2005).

As described previously, the first of these H5N1 HPAI viruses in Asia was identified in 1996 in sick geese from Guangdong Province in southern China (Xu et al, 1999). Related H5N1 HPAI viruses then caused the outbreak in HKSAR in 1997, in which both poultry and humans were affected. This was followed by further outbreaks in HKSAR (in poultry only) from 2001 to early 2002 (Sims et al, 2003a) and poultry and captive wild birds in 2002-03 (Ellis et al, 2004). By this time, infection with a range of Asian-lineage H5N1 HPAI viruses had been detected in an arc extending from Hanoi to Shanghai (Sims et al,2005) and even to provinces further north (see for example A/chicken/Hebei/718/01[H5N1], GenbankTaxonomy ID: 367715). Remarkably, apart from HKSAR, no disease was officially reported in poultry in any of these places (Sims et al, 2005).

From 2003 onwards, infection and disease spread widely to three continents, initially through East and Southeast Asia in 2003-04, and then into southern Russia, the Middle East, Europe, Africa and south Asia in 2005-06.

The precise events that led to broad transborder spread of these viruses are not known, but among the likely contributory factors are the changes that allowed the H5N1 HPAI viruses to multiply subclinically in ducks (see section 3.1), presumably leading to a high baseline level of infection in the region from which the viruses could spread. The presence of high quantities of virus in domestic ducks in a system where live poultry are moved long distances and sold in live-bird markets (Li et al, 2003) without being tested for infection would facilitate spread of infection within, and perhaps in some cases, between countries.

The adaptation to ducks may also have favoured subsequent spread of H5N1 by wild Anatidae (Gilbert et al, 2006a; Sturm-Ramirez et al, 2005). Limited laboratory studies on wild ducks indicate that they can be infected subclinically with Asian-lineage H5N1 HPAI viruses, even after inoculation with very low doses of virus. The clinical outcome depends on a number of factors, including the age of the ducks at the time of inoculation, the species of duck and the strain of virus (Brown et al, 2006). The duration of excretion in most subclinically infected wild ducks was short, but still long enough to allow for transfer from one site to another. There is also considerable circumstantial evidence from Europe, Russia and Mongolia to indicate that wild birds played a significant role in the spread of these viruses (ProMED-mail, 2005; Gilbert et al, 2006b; Irza, 2006) (see section 2.3).

Historically, once established in an area, spread of HPAI was associated with movement of contaminated objects or infected poultry (Alexander, 2000; Wilkinson and Waterson, 1975). This is still considered to be the case today, but the market chains and extent of movement of poultry and products are often more complex than those seen early last century, complicating investigations and analyses. In most cases, epidemiological studies are conducted some time after the disease occurs, leaving investigators the task of determining and ranking probabilities on the basis of imperfect and sometimes complex historical information.

Epidemiological investigations should be conducted on all new outbreaks through tracing studies in an attempt to determine the most likely source of infection. A large number of outbreaks occurring simultaneously over a wide area suggests undetected cases of infection. This occurred in early 2004, when many seemingly disconnected cases of H5N1 HPAI were reported in a number of Asian countries. When such situations arise, further investigations are needed to establish the pathways of infection and the connections between the cases. This type of analysis has not been done for many of the outbreaks of HPAI in Asia and, until such information is available, suggestions on the mode of spread for many of these outbreaks remain speculative.

The recent outbreak of equine influenza in Australia provides some insights into the different routes of transmission of influenza viruses in animal populations. Equine influenza virus had never previously been reported in Australia and therefore horses in Australia (other than a very small percentage of horses vaccinated for international travel) were highly susceptible to infection. The virus was almost certainly introduced to the country via an imported infected stallion, with initial spread probably occurring via fomite transfer. This was exacerbated by subsequent movement of infected horses and by additional fomite transfer. Local short distance spread, possibly aerogenous, also occurred. This epidemic demonstrated that different mechanisms may be involved in the first introduction of virus to those associated with subsequent spread. This has also been proposed in recent investigations of outbreaks of H5N1 HPAI in poultry in eastern Turkey (Bayraktar, 2007), where it is suspected that wild bird introduction was followed by spread through local poultry trade.