Highly pathogenic avian influenza

Extracted from H5N1 highly pathogenic avian influenza worldwide in 2010; FAO EMPRES 

Transboundary Animal Diseases Bulletin, 37: 21-29. 

http://www.fao.org/docrep/014/i2249e/i2249e00.pdf

H5N1 highly pathogenic avian influenza worldwide in 2010

Introduction

A total of 63 countries/territories in Asia, Europe and Africa have been affected by H5N1 highly pathogenic avian influenza (HPAI) since the beginning of the epizootic in poultry in 2003. Following three years of progressive retraction, geographic expansion of the virus was again noted in 2010, with 18 countries affected (14 in Asia, three in Europe, and one in Africa), compared with only 12 countries in 2009 (Table 1 and Figure 1). However, H5N1 HPAI outbreaks still occurred predominantly in Asia, as was also the case in the previous three years. The peak of disease activity was once again seen during the months of January to March (as in Bangladesh, shown in Figure 3). At the country level, the numbers of outbreaks generally decreased or remained at the same level, with the exception of those in Egypt (Table 1), where the number of reported outbreaks has increased, following a change in the disease reporting system. The activities of the Community Animal Health Outreach (CAHO) project, initiated in December 2008, were expanded in February 2009, following an apparent enhancement of outbreak notification. Bhutan was newly infected in 2010, reporting H5N1 HPAI outbreaks in February and March for the first time. Bulgaria, Israel, Japan, Myanmar, the Republic of Korea and Romania had not reported outbreaks of H5N1 HPAI in 2009; the disease was reintroduced into these countries (or reoccurred) in 2010. The reported outbreaks in Romania in March 2010 were the first H5N1 HPAI poultry outbreaks in Europe since October 2008, when the H5N1 virus was detected during routine surveillance in a poultry flock in Germany. It is evident that the threat from H5N1 HPAI remains, including the risk of human infection. In 2010, the World Health Organization (WHO) reported 48 human cases of H5N1 infection – of which 24 were fatal – in five countries: Egypt (29 cases, including 13 fatalities), Indonesia (nine cases, including seven fatalities), Viet Nam (seven cases, including two fatalities), China (two cases, including one fatality) and Cambodia (one fatality). The same countries also reported human cases in 2009. The incidence of the disease in humans decreased in 2010, except for in Viet Nam, where it increased from five to seven cases.

Mixing species, as in this backyard poultry farm, may lead to expansion of the host range of HPAI, and increase the risks of recombination of influenza viruses

Sources: FAO EMPRES-i; OIE WAHID.

Four high-burden countries remain enzootic

As was also the case in 2009, in 2010 the great majority of H5N1 HPAI outbreaks in animals occurred in four high-burden countries, where the disease is deeply entrenched: Bangladesh (2 percent of total cases), Egypt (25 percent), Indonesia (68 percent) and Viet Nam (3 percent) (Table 1). Progressive control of avian influenza (AI) in these settings poses a formidable challenge, and elimination remains the long-term goal. In eastern Asia, the mixing of poultry, pig and human influenza A viruses gives rise to an ever-expanding gene pool of virus subtypes, clades and lineages.  

Bangladesh: During 2010, there were 31 H5N1 HPAI poultry outbreaks reported in Bangladesh (Figure 3), with the majority on commercial farms (29) and just two in backyard holdings. Approximately 175 000 birds were culled. In 2009, there were 32 outbreaks reported, with a high proportion of events observed between January and April in both years. Although no outbreak was observed between June and December 2010, the country is still believed to remain endemic, with active circulation of the virus (clade 2.2). In particular, 2010 virus isolates grouped in sublineage 3 and clustered with sequences of viruses from Bangladesh isolated in 2007 to 2009. This suggests that the virus is maintained in unnoticed reservoirs within the country, possibly domestic ducks. Bangladesh recorded fresh H5N1 HPAI outbreaks in January and February 2011. In total, 49 out of 64 districts recorded outbreaks both on commercial farms and in backyard holdings.

Indonesia continued to report a high number of H5N1 HPAI outbreaks in poultry, as it has done for the past four years. More outbreaks were reported in this country than in the rest of the world combined (Figure 4). H5N1 HPAI is endemic on Java, Sumatra and Sulawesi, with sporadic outbreaks reported elsewhere. Highincidence areas are recognized at both theprovincialand district levels on Java (especially Yogyakarta) and in the south of Sumatra (Lampung). The high number of reports stems in part from the Participatory Disease Surveillance and Response (PDSR) programme (established and supported by FAO), which targets village poultry production systems (mainly backyard) and reports outbreaks at the village level. Only two of Indonesia’s 33 provinces (Maluku and North Maluku) have never reported H5N1 HPAI.

Viet Nam in 2010 reported 48 H5N1 HPAI outbreaks in 20 of its 63 provinces (32 percent), mostly on duck farms (83 percent) and in the small-scale commercial sector: 61 percent of outbreaks were in flocks of 50 to 1 000 birds. Approximately 45 000 birds were culled. A total of 56 outbreaks had been reported in 2009. Active virus circulation surveillance was carriedoutineight target provinces and cities. Prevalences of influenza type A and H5N1 avian viruses were 0.94 and 0.67 percent, respectively, in ducks, and 0.54 and 0 percent in chickens. There was no evidence of AI virus persistence in Muscovy ducks. In 2010 three virus clades were isolated: i) haemagglutinin (HA) clade 1, predominantly in southern Viet Nam and also in Cambodia; ii) HA clade 2.3.4, predominantly in northern Viet Nam during the first half of 2010, and also circulating in China; and iii) HA clade 2.3.2, detected for the first time in late 2009, with four cases detected in the first half of 2010, and others – predominantly in northern Viet Nam, but also isolated in the south – since September 2010. 

Egypt, which reported its first H5N1 HPAI outbreak in February 2006, is considered endemic, regularly reporting outbreaks from almost all of its 29 governorates. During 2010, 443 outbreaks were observed, mainly in backyard systems, against 176 in 2009. However, detection has probably improved because of implementation of the PDSR-like CAHO programme in ten governorates (Bani-Seuif, Behera, Dakahlia, Fayoum, Gharbia, Kafr el Shiekh, Menia, Menufia, Qalubia and Sharkia). Viruses isolated during 2010 were genetically similar to those isolated in 2009. 

Some knowledge gaps remain in the role of wild birds  

During 2010, H5N1 AI events in wild birds continued, with a slight increase from 2009, when 12 were reported. There were 16 events reported in 2010, with Bulgaria, China, its Hong Kong SAR, Indonesia, Mongolia and the Russian Federation each reporting one, Japan reporting four, and the Republic of Korea reporting six (Figure 5). None of the events in 2010 were on the scale of thousands of wild birds dying within weeks, as was the case in 2005 in China and the Russian Federation. The current H5N1 HPAI viruses may be less deadly to wild birds. 

The wild bird species infected in 2010 were mallard, hooded crane, tundra swan, mute swan, eagle owl, 

great crested grebe, goosander, grey heron, gadwall, Eurasian spoonbill, red-billed chough, bar-headed 

goose, brown-headed gull, whooper swan, greylag goose, common buzzard and barn swallow. Since the 

beginning of the epizootic, more than 100 species from 13 orders of birds have been found to be infected with 

H5N1 AI virus. 

Wild bird H5N1 AI outbreaks in 2010 were reported in single birds found during routine surveillance in China, Hong Kong SAR, Indonesia, Bulgaria, Japan and the Republic of Korea. However, this was not the case for events in Mongolia (n = 26 ) and China (n = 170) in May 2010, the Russian Federation (n = 367) in June 2010, and the Republic of Korea (n = 20) in December 2010, where larger numbers of wild birds were found dead. For both Mongolia and the Russian Federation, similar temporal and spatial patterns of wild bird deaths have been observed in past years. Mortalities occur during the northern hemisphere spring season, as birds arrive at breeding grounds.

H5N1 phylogenetic data and evolution of clade 2.3.2 viruses  

H5 HPAI outbreaks worldwide are caused by clade 1 and 2 viruses, particularly clades 2.1, 2.2 and 2.3 for the latter. The following outlines the different clades identified in 2010 (see also Table 2).

Source: FAO EMPRES-i.

All human and animal H5N1 isolates from Cambodia analysed since 2004, including all those from 2010, form part of clade 1 (genotype Z). Recent viruses are similar to the viruses circulating there in previous years. This is the same clade that circulates predominantly in southern Viet Nam, in the Mekong Delta extending into Cambodia. 

Clade 2.1 is known to be circulating in Indonesia, although no 2010 sequence data are available. Data from 2009 show that Indonesian isolates form part of clade 2.1.3, and continue to diversify with the greatest variation currently observed for viruses from Sumatra. 

Clade 2.2 includes 2010 sequences from Nepal, Bangladesh, Bhutan and India. Clade 2.2.1 viruses continue to circulate in Egypt. This clade has also been confirmed for an emu isolate from Israel in September 2010, which suggests some movement of H5N1 HPAI in the region, possibly involving “bridge” species. It is interesting to note that in Egypt the 2.2.1 viruses cluster into two distinct phylogenetic groups: A and B. The majority of the samples belong to group B (or group F according to the United States Centers for Disease Control and Prevention [CDC] nomenclature), while the remainder belong to group A (or C according to CDC), to which most human samples belong. Viruses from group A are predominantly from backyard birds, and those from B from vaccinated commercial poultry. Preliminary studies suggest there is little to no human influenza cross-reactivity between the two groups. Epidemiologically, H5N1 HPAI poultry outbreaks in Egypt continue throughout the Nile Delta, mainly in areas of high human and poultry density.

Clade 2.3.4 sequences were obtained from Viet Nam, Lao People’s Democratic Republic and Myanmar. Clade 2.3.4 isolates from Myanmar are similar to 2007 isolates from the Yangon area, suggesting that the 2010 virus originates from a reservoir in domestic duck flocks. Samples from Lao People’s Democratic Republic identified as clade 2.3.4 cluster with viruses seen in that country during previous years. In late 2010, clade 2.3.2, which was originally detected from a dead Chinese pond heron in China, Hong Kong SAR in 2004, has become the most common clade in the spread of H5N1 HPAI and its invasion into new territories. Bulgaria, China (including Hong Kong SAR), Japan, Mongolia, Myanmar, Nepal, Romania, the Republic of Korea, the Russian Federation and Viet Nam have all been affected by clade 2.3.2 since January 2010. 

In 2010, Nepal was the first country to detect clade 2.3.2 in the South Asia region. Clade 2.3.2 viruses in Nepal were most closely related to viruses isolated in wild birds in 2009 from the Russian Federation, Mongolia and Bulgaria, and also from poultry outbreaks in Romania. The clade 2.3.2 viruses isolated in wild birds in China, Hong Kong SAR and in poultry in Viet Nam were slightly less closely related to these viruses, probably indicating evolution from the source. The isolate from wild birds found dead in June 2010 in the Republic of Tyva, Russian Federation, close to the border with Mongolia, belonged to clade 2.3.2 of the Asian lineage, with 99 percent similarity to earlier 2009/2010 H5N1 isolates from wild birds in Mongolia, Tyva (Russian Federation) and Qinghai (China).  

Most of the recent invasions of clade 2.3.2 may be linked to wild birds. The main clades currently found in wild birds are 2.3.2 and 2.2. The latter has circulated in wild birds since 2005. The predominant clade in 2010 was 2.3.2. For the cases in Bulgaria, China, Hong Kong SAR, Mongolia and the Russian Federation, only wild bird events have been reported, with no poultry affected. The initial clade 2.3.2 incidents in Japan, the Republic of Korea and Romania were also observed in wild birds, and only later did the disease apparently spread to poultry. Other than China, Myanmar and Nepal, all clade 2.3.2-affected countries had been free from the disease for extended periods. 

It is likely that the clade 2.3.2 virus first spilled over from poultry to wild birds and was periodically transported by migratory birds to other locations. Research has demonstrated that some wild waterfowl can shed H5N1 AI virus asymptomatically for two to five days, and migration-disease ecology studies conducted by FAO have established the distances wild waterfowl can migrate over this same time frame. Although individual flights can be long (hundreds of kilometres), wild waterfowl’s transmission of virus over such distances is more likely to occur through a leap-frog migration pattern, with concurrent transmission of virus from one bird to others at stop-over sites, and the newly infected birds carrying virus to the next site.  

Perspectives for 2011 

A number of concerns for 2011 must be considered. The endemic foci of disease will continue to pose a threat to countries at risk in Asia and to Egypt, and a long-term approach is required to eliminate H5N1 HPAI from poultry in these areas. The finding of a new clade of H5N1 (clade 2.3.2) in a wide range of wild birds would suggest that another round of outbreaks in a wider geographic area (Palearctic) may occur, certainly in countries in the eastern parts of Asia. The epidemiology of H5N1 in wild birds is far from understood, and the role of the virus’s evolution in domestic waterfowl as a source of new viruses for wild birds needs to be defined. If H5N1 viruses are also entrenched in wild bird populations, as was found at Qinghai Lake in China, there will be considerable challenges in keeping poultry populations free from disease. It is likely that the trend of increased numbers of outbreaks in early 2011 will be seen again this autumn, and there is likely to be a parallel increase in human cases. 

FAO’s Animal Health Service and its Global Programme for the Prevention and Control of H5N1 HPAI continue to contribute to country preparedness, capacity building and outbreak response in affected countries. Between 2005 and 2010, FAO implemented 170 H5N1 HPAI projects, with 28 still ongoing as of February 2011, the majority in Asia. Veterinary epidemiology teams remain at the strategically placed Regional Animal Health Centres. Regional laboratory and epidemiology networks are also being sustained. A special wildlife unit is dedicated to research into the epidemiology of H5N1 in wild birds and wild birds’ role in the transmission of the virus. 

Contributors: Sophie von Dobschuetz (FAO), Jennifer Siembieda (FAO), 

Mia Kim (FAO), Julio Pinto (FAO) and Scott Newman (FAO)

To top