Tweet
"Outbreaks of HPAI H5N1 virus in Europe during 2005/2006"
v. 1 - June 30, 2006 from Defra, UK
(Lots of good maps, timelines, summaries, etc.)
6 Discussion
6.1 The virus
Genetic analysis of the recent isolates shows that three separate lineages of the Asian strain of the virus have been detected in the EU Member States and neighbouring countries and indicates a level of relationship between different outbreaks in Europe and varying host specificity.
One lineage comprises isolates that have been primarily been isolated from wild birds with no associated outbreaks in poultry. This lineage would indicate direct evidence of linkage between the affected areas in China (Qinghai lake), southern Siberia (Novosibirsk) and in the EU (Germany ? R?gen Island), Sweden and Denmark. The virus isolated from a single dead Whooper swan found washed up in a harbour in Scotland (UK) also belongs to this lineage.
Information from the World Reference Laboratory for Avian Influenza, Weybridge, UK, suggest that the H5N1 virus isolates from Turkey and Romania obtained in October 2005 are almost identical to the virus isolated in wild birds in Central Asia in May 2005. However, genetic analysis of subsequent isolates suggest that this virus may have subsequently evolved into two clades ? one clade is indicative of the lineage of the virus that may be circulating between wild birds and domestic poultry (e.g. Romania). The other clade is indicative of the lineage of the virus that may be circulating between domestic poultry without involvement of wild birds (e.g. Turkey). Matroshovic and others (1999) consider that avian influenza viruses from aquatic birds undergo significant selective pressure in chickens, leading to definite changes in both the HA and the NA during the adaptation process. A single substitution in the HA may result in additional glycosylation sites which in turn may render the virus highly pathogenic in chicken by increasing the release of the virus from cells which facilitates its spread and replication in different tissues of chicken. These features of the HA and NA clearly separate chicken viruses from the viruses of wild aquatic birds.
6.2 The host
6.2.1 Infection
During the period from February to April 2006, H5N1 has been detected in dead wild waterbirds collected at numerous locations in several affected EU Member States. With regard to the species affected, it is notable that most detections of the virus in the affected EU Member States were made in dead wild waterbirds of the order of Anseriformes (swans and ducks). There is, however, emerging circumstantial evidence that live waterbirds (e.g. swans) and some scavenging birds (e.g. gulls) may be infected with HPAI H5N1 virus without showing clinical signs.
It is also notable from the above maps that the virus was detected in greater numbers of dead wild ducks in northern EU Member States compared to a greater number of detections in swans in other affected central and southeastern EU Member States.
Sharp and others (1997) consider that different avian influenza (AI) A virus subtypes are maintained by different avian species. While a particular subtype may infect more than one avian species, they appear to have different, species-specific levels of adaptation.
We have no information on the age and sex of affected wild waterfowl. We assume that all detections were in adult birds. It is however, interesting to note that Sharp and others (1997) consider that juvenile ducks were significantly more likely to be infected with AI type A viruses than adults. It is unknown whether this would apply to infection with HPAI H5N1 virus as well and whether they would be likely to survive such infection and shed the virus for extended period of time.
In experimental conditions, the virus killed seven out of eight 2-week old ducks. Two out of 5-week-old-ducks inoculated with the virus died. The virus did not produce clinical signs in any of these ducks but did infect them. These results confirm that some of the circulating H5N1 isolates are capable of causing disease and death in ducks, however, lethality is age dependant (Pantin-Jackwood and others, 2006). This could suggest that a proportion of ducks infected with the virus could enter the moulting period.
Simulation studies with LPAI virus showed peaks of viral prevalence after nesting due to the population recruitment and during moult period due to high host density. The estimated host population threshold for virus persistence is 380 susceptible individuals on day 1 (Guberti and others, 2006). It remains unknown to what extent this would apply to infection with HPAI H5N1 virus.
6.2.2 Migration
Our previous risk assessment emphasised that caution is required when generalising trends that may relate to carriage of the HPAI H5N1 virus or any other HPAI virus to different regions or countries by migratory birds because they use different migratory routes (flyways).
There may be some limited mixing of the waterbird populations in northern Russia from the four major flyways in Eurasia. However, it is uncertain at this stage whether there is any significant geographic and temporal overlap of these waterbird populations in northern Russia with the waterbird populations in southern Siberia.
Therefore, the level of risk, which will vary from one season of the year to another, will depend on migratory pathways, either direct from infected areas or through contact at intermediate ?mixing? points for migratory species. The evidence to quantify this risk seems to be incomplete. Systematic studies are therefore required to understand these routes, the species susceptibility, pathogenesis and ecology of the virus.
In their official notification to the OIE on 21 October 2005, the Russian authorities confirmed that the H5N1 virus has been confirmed in ducks, muskovy ducks, chicken, geese and turkeys in a number of backyard farms in Tula (Moscow region). This appears to be the only report of the H5N1 virus being detected in western Russia.
In our previous risk assessments we have considered that, should more outbreaks of the H5N1 virus be detected in wider areas of western Russia, this will impact on the likelihood of the virus introduction to the UK. That is, this changed situation could indicate that the virus may be present in migratory populations that arrive to the UK from further east in northern Russia. The expert ornithologists consider that this area would be within the direct migratory routes that exist between the northern Russia and the UK and involve greater numbers of migratory waterfowl.
Around 5 million waterbirds are present in Britain in winter. While some species are resident in the UK (i.e. birds present in winter that have bred here) many species arrive in the UK from arctic areas of North America, Greenland, Iceland, Fenno-Scandia and further east in northern Russia. Many of the waterbird species or populations wintering in the UK derive from northern (arctic or sub-arctic) areas and are highly unlikely to act as carriers of the virus to the UK from the current outbreaks in central Asia. Further, several species of wildfowl have a marine distribution during winter, and, remaining at sea, will therefore not come in to contact with farms or domestic livestock (Cranswick, 2005).
The Volga Basin and North Caspian regions are considered cross-roads for migratory waterbirds that use four major routes in Eurasia and East Europe. These two regions host the vast majority of migratory birds which are nesting in Eastern Fennoscandia, Northern-Central territories of the Russian plain, Ural and parts of western Siberia on their way to overwintering grounds in East Africa. A small proportion of these birds spend winters in south-western Asia. (Lvov and others 2001). Some species of ducks migrate from their breeding grounds in western Siberia to the area around the Caspian Sea.
The existence of complex migratory pathways mean there is a possibility that a small number of individual birds, from a few species, could migrate to Western Europe from areas in Europe currently affected with highly pathogenic AI (HPAI). These represent only a small proportion of individual birds arriving in the UK.
Ringing recoveries (Wernham and others, 2002) show there is some, albeit limited, movement of birds between the UK and southern Russia. Therefore, the inferences about the scale and regularity of the movement of birds between Southern Russia and the UK can only be preliminary and need to be treated with caution. For example, some of the extreme eastern recoveries of UK birds are highly unlikely to have travelled that far in a single winter. Rather, they may have paired with different mates in different breeding seasons and their wintering and breeding grounds may have moved east. Lastly, the UK plays host to only a proportion of these species? populations during the winter: the majority of individuals breeding in Russia are highly unlikely to reach Britain because they winter further east in Europe, for example in the Mediterranean, and for some species in the Baltic (Cranswick, 2005)
It needs to be emphasised that ?although bird banding (ringing) has enabled scientists to gather very detailed information on birds, tracking the birds to understand their movements is a difficult task? (Anonymous, 2005). We have been advised by experts that there is a great amount of ring recovery data. This data is held by individual schemes within the EU member states and centrally at the EU level on behalf of the different schemes for birds ringed in Europe. We consider that it would be useful if this data could be collated and analysed on a continental and national scale to provide much more information on bird movements.
6.3 The environment
6.3.1 Geographic spread
Severe weather conditions in the affected areas in the EU neighbouring countries and the EU may have caused temporary and erratic displacement of unknown numbers of birds from either northern parts of Europe or the Black Sea region, some of which may have been infected with the virus.
7 Conclusions
It is only a relatively short time since the H5N1 strain of the virus was first detected in wild birds. This was in Hong Kong in 2002. Our knowledge of the epidemiology of the virus still remains limited particularly as the study of infection in wild bird populations is difficult. Sufficiently large samples of live birds are impossible to obtain and there is therefore a reliance on the sampling of birds found dead, which involves its own practical difficulties and biases.
Our knowledge of the epidemiology of H5N1 in wild birds Europe is limited as it is only a few months since the infection was first detected. Understanding the current epidemiological picture and predicting the future occurrence has limitations and any inferences made therefore have a great degree of uncertainty. However, the available information is clearly of value in starting to improve our ability to assess the risks. The following indicates the key findings and observations to date.
The virus isolates in the EU and the EU neighbouring countries appear to be genetically closely related to the Asian lineage of the virus that has been isolated in China (Qinghai Lake), Russia (Southern Siberia) and Mongolia.
They are distinguishable from the apparently chicken-adapted strains infecting domestic poultry in Turkey.
On the basis of the presented limited information, it would appear that multiple introductions of the virus in Europe resulting in three genetic groupings may have occurred during the winter 2005. The initial outbreaks were sporadic and occurred at different and distant geographic locations within a relatively short period of time. Subsequent outbreaks peaked in March 2006 and became clustered geographically and in time in some EU Member States. These outbreaks may have coincided with wild bird population displacements and their congregation at limited habitats due to exceptional environmental conditions.
On the basis of genetic studies, migratory wild waterfowl may have had a role in the virus introduction and subsequent spread in Europe. The HPAI H5N1 virus was primarily detected in two species of dead wild waterfowl (swans and ducks) that belong to the Order Anseriformes. These species appear to play a greater role in the epidemiology of the virus. This finding is not unexpected as ducks, geese, swans and related web-footed birds are recognised as the primary reservoir hosts for influenza A viruses.
There is growing evidence that the virus has been detected in live and apparently healthy wild waterfowl (i.e. swans and ducks and scavenging birds (i.e. gulls). Given that spring migration of wild waterfowl may have been largely completed, these developments could suggest that the virus may have become established in local wild bird population in some EU Member States. One indication of such development could be further sporadic detections from these Member States, particularly during the forthcoming moulting period when the resident wild waterfowl will congregate in large numbers.
Other wild bird species (e.g. raptors, other wild birds) could be affected with a fatal outcome but appear to play a lesser role in the epidemiology of the virus. However, more systematic studies at various levels (i.e. local, national, international) involving different agencies and interest groups are required to ascertain their role as potential carriers of the virus without showing clinical signs of the disease.
Wild bird migration is a natural phenomenon that cannot be controlled, therefore, the likelihood of the virus introductions to the EU during the forthcoming migration along with other potential pathways would have to be re-assessed taking into account epidemiological developments.
There are a number of aspects of the epidemiology of infection in wild birds which are important in assessing the risks from the coming migration. These include the maintenance of infection in breeding grounds in northern Russia, and if infection persists whether infection will be more widespread in this area, particularly the more western parts.
It is unknown at present whether H5N1 infection will persist in wild bird populations throughout the year in Europe in the absence of further introductions. Similarly, if infection does persist, there is uncertainty as to whether infection will extend geographically and/or become established in a wider range of wild bird species. The continued surveillance is crucial in this respect, as is the analysis of the accumulated data from the EU Member States, at least.
The current situation in Europe suggests that extensive surveillance, complemented with appropriate biosecurity measures were an effective way of detecting and preventing the introduction of the virus into commercial poultry operations. However, epidemiological studies are required in the EU Member States in the areas where infection has been found in wild birds to identify the domestic poultry flocks that could be regarded as at risk. This would provide valuable epidemiological information to support the apparent effectiveness of the biosecurity measures. Obtaining epidemiological evidence on this aspect is particularly important.
The identification and understanding of the risk factors for the various species of farmed poultry, kept under the range of management systems used in the EU countries is also important, as should infection become established in a wild birds there are no acceptable means of control in such free living populations. Preventing the transfer of infection to domestic poultry, which if it occurred would increase the risk for infection of the human population, is therefore paramount.
v. 1 - June 30, 2006 from Defra, UK
(Lots of good maps, timelines, summaries, etc.)
6 Discussion
6.1 The virus
Genetic analysis of the recent isolates shows that three separate lineages of the Asian strain of the virus have been detected in the EU Member States and neighbouring countries and indicates a level of relationship between different outbreaks in Europe and varying host specificity.
One lineage comprises isolates that have been primarily been isolated from wild birds with no associated outbreaks in poultry. This lineage would indicate direct evidence of linkage between the affected areas in China (Qinghai lake), southern Siberia (Novosibirsk) and in the EU (Germany ? R?gen Island), Sweden and Denmark. The virus isolated from a single dead Whooper swan found washed up in a harbour in Scotland (UK) also belongs to this lineage.
Information from the World Reference Laboratory for Avian Influenza, Weybridge, UK, suggest that the H5N1 virus isolates from Turkey and Romania obtained in October 2005 are almost identical to the virus isolated in wild birds in Central Asia in May 2005. However, genetic analysis of subsequent isolates suggest that this virus may have subsequently evolved into two clades ? one clade is indicative of the lineage of the virus that may be circulating between wild birds and domestic poultry (e.g. Romania). The other clade is indicative of the lineage of the virus that may be circulating between domestic poultry without involvement of wild birds (e.g. Turkey). Matroshovic and others (1999) consider that avian influenza viruses from aquatic birds undergo significant selective pressure in chickens, leading to definite changes in both the HA and the NA during the adaptation process. A single substitution in the HA may result in additional glycosylation sites which in turn may render the virus highly pathogenic in chicken by increasing the release of the virus from cells which facilitates its spread and replication in different tissues of chicken. These features of the HA and NA clearly separate chicken viruses from the viruses of wild aquatic birds.
6.2 The host
6.2.1 Infection
During the period from February to April 2006, H5N1 has been detected in dead wild waterbirds collected at numerous locations in several affected EU Member States. With regard to the species affected, it is notable that most detections of the virus in the affected EU Member States were made in dead wild waterbirds of the order of Anseriformes (swans and ducks). There is, however, emerging circumstantial evidence that live waterbirds (e.g. swans) and some scavenging birds (e.g. gulls) may be infected with HPAI H5N1 virus without showing clinical signs.
It is also notable from the above maps that the virus was detected in greater numbers of dead wild ducks in northern EU Member States compared to a greater number of detections in swans in other affected central and southeastern EU Member States.
Sharp and others (1997) consider that different avian influenza (AI) A virus subtypes are maintained by different avian species. While a particular subtype may infect more than one avian species, they appear to have different, species-specific levels of adaptation.
We have no information on the age and sex of affected wild waterfowl. We assume that all detections were in adult birds. It is however, interesting to note that Sharp and others (1997) consider that juvenile ducks were significantly more likely to be infected with AI type A viruses than adults. It is unknown whether this would apply to infection with HPAI H5N1 virus as well and whether they would be likely to survive such infection and shed the virus for extended period of time.
In experimental conditions, the virus killed seven out of eight 2-week old ducks. Two out of 5-week-old-ducks inoculated with the virus died. The virus did not produce clinical signs in any of these ducks but did infect them. These results confirm that some of the circulating H5N1 isolates are capable of causing disease and death in ducks, however, lethality is age dependant (Pantin-Jackwood and others, 2006). This could suggest that a proportion of ducks infected with the virus could enter the moulting period.
Simulation studies with LPAI virus showed peaks of viral prevalence after nesting due to the population recruitment and during moult period due to high host density. The estimated host population threshold for virus persistence is 380 susceptible individuals on day 1 (Guberti and others, 2006). It remains unknown to what extent this would apply to infection with HPAI H5N1 virus.
6.2.2 Migration
Our previous risk assessment emphasised that caution is required when generalising trends that may relate to carriage of the HPAI H5N1 virus or any other HPAI virus to different regions or countries by migratory birds because they use different migratory routes (flyways).
There may be some limited mixing of the waterbird populations in northern Russia from the four major flyways in Eurasia. However, it is uncertain at this stage whether there is any significant geographic and temporal overlap of these waterbird populations in northern Russia with the waterbird populations in southern Siberia.
Therefore, the level of risk, which will vary from one season of the year to another, will depend on migratory pathways, either direct from infected areas or through contact at intermediate ?mixing? points for migratory species. The evidence to quantify this risk seems to be incomplete. Systematic studies are therefore required to understand these routes, the species susceptibility, pathogenesis and ecology of the virus.
In their official notification to the OIE on 21 October 2005, the Russian authorities confirmed that the H5N1 virus has been confirmed in ducks, muskovy ducks, chicken, geese and turkeys in a number of backyard farms in Tula (Moscow region). This appears to be the only report of the H5N1 virus being detected in western Russia.
In our previous risk assessments we have considered that, should more outbreaks of the H5N1 virus be detected in wider areas of western Russia, this will impact on the likelihood of the virus introduction to the UK. That is, this changed situation could indicate that the virus may be present in migratory populations that arrive to the UK from further east in northern Russia. The expert ornithologists consider that this area would be within the direct migratory routes that exist between the northern Russia and the UK and involve greater numbers of migratory waterfowl.
Around 5 million waterbirds are present in Britain in winter. While some species are resident in the UK (i.e. birds present in winter that have bred here) many species arrive in the UK from arctic areas of North America, Greenland, Iceland, Fenno-Scandia and further east in northern Russia. Many of the waterbird species or populations wintering in the UK derive from northern (arctic or sub-arctic) areas and are highly unlikely to act as carriers of the virus to the UK from the current outbreaks in central Asia. Further, several species of wildfowl have a marine distribution during winter, and, remaining at sea, will therefore not come in to contact with farms or domestic livestock (Cranswick, 2005).
The Volga Basin and North Caspian regions are considered cross-roads for migratory waterbirds that use four major routes in Eurasia and East Europe. These two regions host the vast majority of migratory birds which are nesting in Eastern Fennoscandia, Northern-Central territories of the Russian plain, Ural and parts of western Siberia on their way to overwintering grounds in East Africa. A small proportion of these birds spend winters in south-western Asia. (Lvov and others 2001). Some species of ducks migrate from their breeding grounds in western Siberia to the area around the Caspian Sea.
The existence of complex migratory pathways mean there is a possibility that a small number of individual birds, from a few species, could migrate to Western Europe from areas in Europe currently affected with highly pathogenic AI (HPAI). These represent only a small proportion of individual birds arriving in the UK.
Ringing recoveries (Wernham and others, 2002) show there is some, albeit limited, movement of birds between the UK and southern Russia. Therefore, the inferences about the scale and regularity of the movement of birds between Southern Russia and the UK can only be preliminary and need to be treated with caution. For example, some of the extreme eastern recoveries of UK birds are highly unlikely to have travelled that far in a single winter. Rather, they may have paired with different mates in different breeding seasons and their wintering and breeding grounds may have moved east. Lastly, the UK plays host to only a proportion of these species? populations during the winter: the majority of individuals breeding in Russia are highly unlikely to reach Britain because they winter further east in Europe, for example in the Mediterranean, and for some species in the Baltic (Cranswick, 2005)
It needs to be emphasised that ?although bird banding (ringing) has enabled scientists to gather very detailed information on birds, tracking the birds to understand their movements is a difficult task? (Anonymous, 2005). We have been advised by experts that there is a great amount of ring recovery data. This data is held by individual schemes within the EU member states and centrally at the EU level on behalf of the different schemes for birds ringed in Europe. We consider that it would be useful if this data could be collated and analysed on a continental and national scale to provide much more information on bird movements.
6.3 The environment
6.3.1 Geographic spread
Severe weather conditions in the affected areas in the EU neighbouring countries and the EU may have caused temporary and erratic displacement of unknown numbers of birds from either northern parts of Europe or the Black Sea region, some of which may have been infected with the virus.
7 Conclusions
It is only a relatively short time since the H5N1 strain of the virus was first detected in wild birds. This was in Hong Kong in 2002. Our knowledge of the epidemiology of the virus still remains limited particularly as the study of infection in wild bird populations is difficult. Sufficiently large samples of live birds are impossible to obtain and there is therefore a reliance on the sampling of birds found dead, which involves its own practical difficulties and biases.
Our knowledge of the epidemiology of H5N1 in wild birds Europe is limited as it is only a few months since the infection was first detected. Understanding the current epidemiological picture and predicting the future occurrence has limitations and any inferences made therefore have a great degree of uncertainty. However, the available information is clearly of value in starting to improve our ability to assess the risks. The following indicates the key findings and observations to date.
The virus isolates in the EU and the EU neighbouring countries appear to be genetically closely related to the Asian lineage of the virus that has been isolated in China (Qinghai Lake), Russia (Southern Siberia) and Mongolia.
They are distinguishable from the apparently chicken-adapted strains infecting domestic poultry in Turkey.
On the basis of the presented limited information, it would appear that multiple introductions of the virus in Europe resulting in three genetic groupings may have occurred during the winter 2005. The initial outbreaks were sporadic and occurred at different and distant geographic locations within a relatively short period of time. Subsequent outbreaks peaked in March 2006 and became clustered geographically and in time in some EU Member States. These outbreaks may have coincided with wild bird population displacements and their congregation at limited habitats due to exceptional environmental conditions.
On the basis of genetic studies, migratory wild waterfowl may have had a role in the virus introduction and subsequent spread in Europe. The HPAI H5N1 virus was primarily detected in two species of dead wild waterfowl (swans and ducks) that belong to the Order Anseriformes. These species appear to play a greater role in the epidemiology of the virus. This finding is not unexpected as ducks, geese, swans and related web-footed birds are recognised as the primary reservoir hosts for influenza A viruses.
There is growing evidence that the virus has been detected in live and apparently healthy wild waterfowl (i.e. swans and ducks and scavenging birds (i.e. gulls). Given that spring migration of wild waterfowl may have been largely completed, these developments could suggest that the virus may have become established in local wild bird population in some EU Member States. One indication of such development could be further sporadic detections from these Member States, particularly during the forthcoming moulting period when the resident wild waterfowl will congregate in large numbers.
Other wild bird species (e.g. raptors, other wild birds) could be affected with a fatal outcome but appear to play a lesser role in the epidemiology of the virus. However, more systematic studies at various levels (i.e. local, national, international) involving different agencies and interest groups are required to ascertain their role as potential carriers of the virus without showing clinical signs of the disease.
Wild bird migration is a natural phenomenon that cannot be controlled, therefore, the likelihood of the virus introductions to the EU during the forthcoming migration along with other potential pathways would have to be re-assessed taking into account epidemiological developments.
There are a number of aspects of the epidemiology of infection in wild birds which are important in assessing the risks from the coming migration. These include the maintenance of infection in breeding grounds in northern Russia, and if infection persists whether infection will be more widespread in this area, particularly the more western parts.
It is unknown at present whether H5N1 infection will persist in wild bird populations throughout the year in Europe in the absence of further introductions. Similarly, if infection does persist, there is uncertainty as to whether infection will extend geographically and/or become established in a wider range of wild bird species. The continued surveillance is crucial in this respect, as is the analysis of the accumulated data from the EU Member States, at least.
The current situation in Europe suggests that extensive surveillance, complemented with appropriate biosecurity measures were an effective way of detecting and preventing the introduction of the virus into commercial poultry operations. However, epidemiological studies are required in the EU Member States in the areas where infection has been found in wild birds to identify the domestic poultry flocks that could be regarded as at risk. This would provide valuable epidemiological information to support the apparent effectiveness of the biosecurity measures. Obtaining epidemiological evidence on this aspect is particularly important.
The identification and understanding of the risk factors for the various species of farmed poultry, kept under the range of management systems used in the EU countries is also important, as should infection become established in a wild birds there are no acceptable means of control in such free living populations. Preventing the transfer of infection to domestic poultry, which if it occurred would increase the risk for infection of the human population, is therefore paramount.