The ongoing animal and human health crises caused by influenza viruses of H5N1 subtype have focused the attention of international organizations and donors on the need for improved veterinary infrastructure in developing countries and the need for improved communication between the human and animal health sectors. The circulation and re-emergence of high-pathogenicity avian influenza viruses of H5N1 subtype are still major concerns because of potential effects on human health, on the profitability of poultry industries, and on the livelihood of the rural environment. Significant improvements toward the management of these outbreaks have occurred worldwide, including new legislative tools, intervention strategies, and investments in capacity building in both developed and developing countries. This has led to a greater understanding of certain aspects of this infection and of its pandemic potential, although we are still far from certainties and from resolving the situation. Given that genetic analysis of the viruses causing human pandemics since the beginning of the 20th century have indicated that at least the hemagglutinin gene was donated from an avian progenitor virus, it would seem reasonable to exploit the information we have from an animal health perspective to support public health policies. Possibly the biggest challenge we have is to find novel ways to maximize the use of the information that is generated as a result of the improved networking and diagnostic capacities. In the era of globalization, emerging and re-emerging diseases of public health relevance are a concern to developing and developed countries and are a real threat because of the interdependence of the global economy. Communication and analysis systems currently available should be tailored to meet global health priorities, and used to develop and constantly improve novel systems for the exploitation of information to generate knowledge. Another fundamental task the veterinary community needs to deliver on is that of bringing relevant information to international discussion tables at which international control and prevention are presented and optimized. The veterinary community has knowledge and areas of expertise that should undoubtedly be part of strategic decisions and are essential to manage the human and animal health implications of avian influenza infections.

Between 2006 and 2008, only one outbreak of highly pathogenic notifiable avian influenza (AI) was reported from the Americas, the Caribbean, and Australia. The outbreak, caused by H7N3, occurred in September 2007 in a multiage broiler breeder facility (∼49,000 birds) near Regina Beach in southern Saskatchewan, Canada. The disease was confined to a single farm; the farm was depopulated. All other reports of infections in poultry or wild birds involved low pathogenicity AI viruses. A notable event that occurred during the 3-yr period was the spread of low pathogenicity notifiable AI (LPNAI) H5N2 (Mexican lineage) into the Caribbean countries of the Dominican Republic and Haiti in 2007 and 2008, respectively, representing the first detection of AI reported in these countries. Mexico reported that the LPNAI H5N2 virus continued to circulate in the central regions of the country, and a total of 49 isolations were made from 12 states between 2006 and 2008. Also, during this period there was a significant increase in AI surveillance in many countries throughout the Americas, the Caribbean, and Australia, resulting in the detection of AI subtypes H1 through H12 and N1 through N9 in domestic bird species (chickens, turkeys, guinea fowl, upland game birds, and ducks/geese). The United States was the only one of these countries that reported detections of LPNAI (H5 or H7) infections in commercial poultry: one in chickens (H7N3, 2007), two in turkeys (H5N1 and H5N2, 2007), and one in pheasants (H5N8, 2008). Detections of AI viruses in wild birds between 2006 and 2008 were reported from North America (Canada and the United States), South America (Bolivia, Argentina, Chile, and Brazil), and Australia.

Events during the period extending from 2006 to 2009 have been overshadowed by the ongoing panzootic with H5N1 (highly pathogenic notifiable avian influenza [HPNAI]), which has afflicted 63 countries and three continents (Africa, Asia, and Europe) during the review period. Two countries, Indonesia and Egypt, have formally declared the disease endemic to the World Organisation for Animal Health, while others have used a variety of approaches aimed at containment, control, and eradication. These approaches have achieved variable success, but in 2009 several countries that had previously declared themselves free of HPNAI became reinfected. In addition, the virus continued to be detected widely in wild bird populations, even in the absence of local poultry outbreaks. Other poultry outbreaks with HPNAI have been reported in South Africa (in ostriches with H5N2 in 2006) and the U.K. (in chickens with H7N7 in 2008). Also notable was the report of H5N2 HPNAI in wild bird populations in North Africa in 2007. Improved active surveillance systems and vigilance for notifiable avian influenza (NAI) in domestic poultry, especially in host groupings, in which clinical signs following infection may be inapparent (e.g., domestic waterfowl), have inevitably resulted in the detection and reporting of other activity. Low pathogenicity NAI H5 or H7 viruses were isolated/detected from poultry in Belgium (H5N2, 2008), Chinese Taipei (H5N2, 2008), Denmark (H5N2, 2006; H7N1, 2008), France (H5N2, 2007), Germany (H7N3, 2008), Italy (H7N7, 2006; H7N3, 2007–08), the Netherlands (H7N7, 2006), Portugal (H5N2, 2007; H5N3, 2007), the Republic of Korea (H7N8, 2007; H5N2, 2008), and the U.K. (H7N3, 2006; H7N2, 2007). In addition, there has also been significant activity with H6 and H9 viruses in poultry populations, especially in Asia.

Since 2005 there have been five incursions into Great Britain of highly pathogenic avian influenza (HPAI) viruses of subtype H5N1 related to the ongoing global epizootic. The first incursion occurred in October 2005 in birds held in quarantine after importation from Taiwan. Two incursions related to wild birds: one involved a single dead whooper swan found in March 2006 in the sea off the east coast of Scotland, and the other involved 10 mute swans and a Canada goose found dead over the period extending from late December 2007 to late February 2008 on or close to a swannery on the south coast of England. The other two outbreaks occurred in commercial poultry in January 2007 and November 2007, both in the county of Suffolk. The first of these poultry outbreaks occurred on a large turkey farm, and there was no further spread. The second outbreak occurred on a free-range farm rearing turkeys, ducks, and geese and spread to birds on a second turkey farm that was culled as a dangerous contact. Viruses isolated from these five outbreaks were confirmed to be Asian H5N1 HPAI viruses; the quarantine outbreak was attributed to a clade 2.3 virus and the other four to clade 2.2 viruses. This article describes the outbreaks, their control, and the possible origins of the responsible viruses.

This paper reviews outbreaks of Asian-lineage highly pathogenic avian influenza virus (HPAIV) H5N1 in wild birds since June 2006, surveillance strategies, and research on virus epidemiology in wild birds to summarize advances in understanding the role of wild birds in the spread of HPAIV H5N1 and the risk that infected wild birds pose for the poultry industry and for public health. Surveillance of apparently healthy wild birds (“active” surveillance) has not provided early warning of likely infection for the poultry industry, whereas searches for and reports of dead birds (“passive” surveillance) have provided evidence of environmental presence of the virus, but not necessarily its source. Most outbreaks in wild birds have occurred during periods when they are experiencing environmental, physiologic, and possibly psychological stress, including adverse winter weather and molt, but not, apparently, long-distance migration. Examination of carcasses of infected birds and experimental challenge with strains of HPAIV H5N1 have provided insight into the course of infection, the extent of virus shedding, and the relative importance of cloacal vs. oropharyngeal excretion. Satellite telemetry of migrating birds is now providing data on the routes taken by individual birds, their speed of migration, and the duration of stopovers. It is still not clear how virus shedding during the apparently clinically silent phase of infection relates to the distance travelled by infected birds. Mounting an immune response and undertaking strenuous exercise associated with long migratory flights may be competitive. This is an area where further research should be directed in order to discover whether wild birds infected with HPAIV H5N1 are able or willing to embark on migration.

In September 2007, an H7N3 highly pathogenic avian influenza outbreak (HPAI) occurred on a multiple-age broiler breeder operation near Regina Beach, Saskatchewan, Canada. Mortality was initially observed in a barn that housed 24-wk-old roosters, with later involvement of 32-wk-old breeders. All birds on the affected premises were destroyed, and surveillance of surrounding farms demonstrated no further spread. The use of water from a dugout pond during periods of high demand, and the proximity of the farm to Last Mountain Lake, the northern end of which is a bird sanctuary, implicated wild aquatic birds as a possible source of the virus. Of particular note, the H7-specific real-time reverse transcription polymerase chain reaction assay that was in use at the time did not detect the virus associated with this outbreak. A Canadian national influenza A virus survey of wild aquatic birds detected no H7 subtype viruses in 2005 and 2006; however, H7 subtype viruses were detected in the fall of 2007. Phylogenetic analysis of a number of these H7 isolates demonstrated an evolutionary relationship with each other, as well as with the H7N3 HPAI virus that was isolated from the Saskatchewan broiler breeder farm.

Human influenza A viruses are classic examples of antigenically variable pathogens that have a seemingly endless capacity to evade the host's immune response. The viral hemagglutinin (HA) and neuraminidase (NA) proteins are the main targets of our antibody response to combat infections. HA and NA continuously change to escape from humoral immunity, a process known as antigenic drift. As a result of antigenic drift, the human influenza vaccine is updated frequently. The World Health Organization (WHO) coordinates a global influenza surveillance network that, by the hemagglutination inhibition (HI) assay, routinely characterizes the antigenic properties of circulating strains in order to select new seed viruses for such vaccine updates. To facilitate a quantitative interpretation and easy visualization of HI data, a new computational technique called “antigenic cartography” was developed. Since its development, antigenic cartography has been applied routinely to assist the WHO with influenza surveillance activities. Until recently, antigenic variation was not considered a serious issue with influenza vaccines for poultry. However, because of the diversification of the Asian H5N1 lineage since 1996 into multiple genetic clades and subclades, and because of the long-term use of poultry vaccines against H5 in some parts of the world, this issue needs to be re-addressed. The antigenic properties of panels of avian H5N1 viruses were characterized by HI assay, using mammalian or avian antisera, and analyzed using antigenic cartography methods. These analyses revealed antigenic differences between circulating H5N1 viruses and the H5 viruses used in poultry vaccines. Considerable antigenic variation was also observed within and between H5N1 clades. These observations have important implications for the efficacy and long-term use of poultry vaccines.

Protective immunity to avian influenza (AI) virus can be elicited in chickens by in ovo or intramuscular vaccination with replication-competent adenovirus (RCA)-free human recombinant adenovirus serotype 5 (Ad5) encoding AI virus H5 (AdTW68.H5) or H7 (AdCN94.H7) hemagglutinins. We evaluated bivalent in ovo vaccination with AdTW68.H5 and AdCN94.H7 and determined that vaccinated chickens developed robust hemagglutination inhibition (HI) antibody levels to both H5 and H7 AI strains. Additionally, we evaluated immune responses of 1-day-old chickens vaccinated via spray with AdCN94.H7. These birds showed increased immunoglobulin A responses in lachrymal fluids and increased interleukin-6 expression in Harderian gland–derived lymphocytes. However, specific HI antibodies were not detected in the sera of these birds. Because pigs might play a role as a “mixing vessel” for the generation of pandemic influenza viruses we explored the use of RCA-free adenovirus technology to immunize pigs against AI virus. Weanling piglets vaccinated intramuscularly with a single dose of RCA-free AdTW68.H5 developed strong systemic antibody responses 3 wk postvaccination. Intranasal application of AdTW68.H5 in piglets resulted in reduced vaccine coverage, i.e., 33% of pigs (2/6) developed an antibody response, but serum antibody levels in those successfully immunized animals were similar to intramuscularly vaccinated animals.

Fowlpox (FP)-vectored avian influenza (FP-AI) vaccines are used in 1-day-old chickens, but they have also recently been shown to be immunogenic in ducks. The objectives of this work were 1) to evaluate safety and to compare the immunogenicity in ducks of three poxvirus vectors (fowlpox, canarypox, and vaccinia) expressing the same hemagglutinin gene from an H5N1 isolate, 2) to study the effect of the dose of the FP-AI and the presence of an adjuvant in 1-day-old Pekin ducks on antibody response after a boost with inactivated vaccine given 3 wk later, and 3) to confirm the immunogenicity of such a heterologous prime-boost vaccination scheme in 1-day-old Muscovy ducks. Immunogenicity induced by the three poxvirus vectors was comparable, and the FP vector was selected for the other studies. As published previously, there was a strong dose effect of the FP-AI priming on the hemagglutination inhibition (HI) titers induced after the boost with an inactivated vaccine. In contrast, the two tested adjuvants did not significantly increase the activity of FP-AI priming. The heterologous prime-boost regimen given to both Muscovy and Pekin ducklings at 1 and 14 or 21 days of age, respectively, was shown to be at least as immunogenic as two administrations of inactivated vaccines given at 2 and 5 wk of age. However, HI antibody titers were of short duration for both vaccine schemes, and their persistence was heterogeneous among individual birds.

The protective dose of a live recombinant LaSota Newcastle disease virus (NDV)–avian influenza H5 vaccine (rNDV-LS/AI-H5) was determined in broiler chickens with high levels of maternal antibodies against NDV and avian influenza virus (AIV). At hatch the geometric mean titers (GMT) of the chickens' maternal antibodies were 2^5.1 and 2^10.3 for NDV and AIV, respectively. At the time of vaccination the GMT was 2^3.1 for NDV and 2^7.9 for AIV. The chickens were vaccinated with one drop (0.03 ml) in the eye at 10 days of age as is typical under field conditions. The test chickens received 10^4.8, 10^5.8, 10^6.8, or 10^7.8 mean chicken embryo infective doses (CEID₅₀) of the rNDV-LS/AI-H5 vaccine. Control chickens were either nonvaccinated, or vaccinated with 10^5.8 or 10^6.8 CEID₅₀ of a commercial live LaSota NDV vaccine. Birds were challenged with either the Mexican highly pathogenic avian influenza virus (HPAIV) strain A/Chicken/Queretaro/14588-19/95 (H5N2) or a Mexican velogenic viscerotropic (VV) NDV strain. One hundred percent of the chickens vaccinated with the rNDV-LS/AI-H5 vaccine were protected against HPAIV and VVNDV when a challenge dose of 10^6.8 EID₅₀ or higher was administered by eye drop. Birds vaccinated with the LaSota NDV vaccine were protected against VVNDV, but not against HPAIV.

Specific-pathogen-free chickens immunized at 14 days of age with either an inactivated recombinant Newcastle disease virus–LaSota/avian influenza H5 (K-rNDV-LS/AI-H5) vaccine or a killed Newcastle disease/avian influenza whole-virus vaccine (K-ND/AI) were protected from disease when challenged with either A/chicken/Queretaro/14588-19/95 (H5N2), a high pathogenicity avian influenza virus (HPAIV) strain isolated in Mexico in 1995, or with a Mexican velogenic viscerotropic Newcastle disease virus (VVNDV) strain 21 days postvaccination. All nonvaccinated chickens challenged with HPAIV or VVNDV succumbed to disease, while those vaccinated with K-rNDV-LS/AI-H5 or K-ND/AI were protected from severe clinical signs and death. Both vaccines induced hemagglutination-inhibition (HI) antibody responses against NDV and AIV. Antibodies against AIV nucleoprotein were not detected by enzyme-linked immunosorbent assay (ELISA) in birds vaccinated with the inactivated rNDV-LS/AI-H5 vaccine. These chickens became positive for AIV antibodies by ELISA only after challenge with HPAIV. The data clearly indicate that the inactivated rNDV-LS/AI-H5 vaccine confers protection comparable to that of the conventional killed whole-virus vaccine against both NDV and AIV, while still allowing differentiation of infected from vaccinated animals by HI and ELISA tests.

Systematic vaccination can be applied when a disease has become enzootic in a country or region. The final goal of the approach is to control or eradicate the disease within the country. This is a long-term vaccination plan that could be applied nationwide to all commercial and backyard poultry. However, after several months of vaccination in enzootic areas, maternally derived antibody (MDA) is present in young chicks, providing some protection and/or interference with vaccination. The aim of this study was to evaluate the level of protection afforded by MDA against challenge with highly pathogenic avian influenza virus (HPAIV), and its suspected interference with current inactivated vaccines in broilers under controlled laboratory conditions. In the first set of experiments, broilers were vaccinated with inactivated vaccines containing H5N2 subtype antigens in the presence or absence of homologue MDAs and challenged with a clade 2.2 H5N1 HPAIV. In the second set of experiments, day-old broilers, either with or without avian influenza MDA, received a regular-type monovalent H5N2 AI vaccine (0.5 ml) or a concentrated (0.2 ml) AI-Newcastle disease virus combined inactivated vaccine subcutaneously. They were then challenged at 11 or 35 days of age. In conclusion, our results indicate that protection induced by day-old administration of inactivated vaccine (regular or concentrated) in the presence or absence of MDA to H5N2 AIV induces poor protection against challenge with H5N1 HPAIV and should not be recommended. Based on our results, vaccination of MDA-positive chickens at a later age (10 days) seems to be a valuable recommendation, although MDAs may still interfere with vaccination to a lesser extent because they are present up to 3 wk posthatch. Therefore, in areas with high infection pressure, when possible, two vaccinations are recommended for optimal protection. Also, it might be advisable to take into account day-old AI MDA titers when one is determining the optimal age of vaccination.

This paper analyzes the efficacy of vaccination to control low pathogenicity avian influenza outbreaks using information collected during four epidemics occurring in Italy between 2000 and 2005. Different vaccination strategies and protocols for meat-turkey immunization are also considered.

The objective of the study was to compare efficacy of two fowlpox (FP) vector vaccines (FP-AI) against H5N1 highly pathogenic avian influenza (HPAI): one (vFP89) expressing the native hemagglutinin (HA) gene from H5N8 A/turkey/Ireland/1378/83 and the other (vFP2211) expressing a modified synthetic HA gene from H5N1 A/chicken/Indonesia/7/2003. Four groups of 20 1-day-old specific-pathogen-free chickens were made: Groups 1 and 2 were immunized with 3 log₁₀ tissue-culture infectious dose 50% (TCID₅₀) of vFP89 and vFP2211, respectively, whereas group 3 was immunized with vFP89, but received a booster immunization at 2 wk of age with an inactivated vaccine containing A/turkey/Wisconsin/68 H5N9 virus (inH5N9); group 4 was left unvaccinated. Ten birds from each group were challenged on day 21 with A/turkey/Turkey/1/2005 clade 2.2 H5N1 HPAI virus. The 10 other chickens from each group were put in contact with their groupmates on day 22. FP-AI induced low hemagglutination inhibition (HI) titers before challenge (GMT <4 log₂) and an HI titer boost was observed 1 wk after the inH5N9 boost. All directly challenged and 9/10 nonvaccinated contact chickens died after challenge (mean death time of 2.3 and 6.1 days, respectively) and most of them shed virus before death via cloacal and buccal routes. All vaccinated birds were clinically protected from HPAI challenge. One (vFP2211), 2 (vFP89 inact.), or 3 (vFP89) out of the 10 directly challenged vaccinated chickens shed virus via the buccal route 2–5 days postinfection. No shedding was detected in the contact-challenged vaccinated birds. Altogether, these data show excellent levels of protection in all three vaccinated groups, and therefore no detectable effect of the origin of the inserted H5 gene on protection under these tested conditions.

Highly pathogenic (HP) H5N1 avian influenza (AI) viruses continue to circulate in Asia and have spread to other regions of the world. Though attempts at eradication of the viruses during various outbreaks have been successful for short periods of time, new strains of H5N1 viruses continue to emerge and have become endemic in parts of Asia and Africa. Vaccination has been employed in Vietnam as part of AI control programs. Domestic ducks, which make up a large part of poultry in Vietnam, have been recognized as one of the primary factors in the spread of AI in this country. As a result, ducks have been included in the vaccination programs. Despite the effort to control AI in Vietnam, eradication of the disease has not been possible, due in part to the emergence and spread of new viruses. Here, we tested the abilities of avian influenza oil emulsion vaccines of different genetic origins to protect against disease and viral shedding in both 2-wk-old white leghorn chickens and 1-wk-old Pekin ducks. Seventy-five to 100% of vaccinated chickens were protected from mortality, but viral shedding occurred for at least 4 days post challenge. All but one vaccinated duck were protected from mortality; however, all groups shed virus up through at least 5 days postchallenge, depending on the vaccine and challenge virus used. Differences in levels of hemagglutination inhibition (HI) antibody titers induced by the vaccines were observed in both chickens and ducks. Although the vaccines tested were effective in protecting against disease and mortality, updated and more efficacious vaccines are likely needed to maintain optimal protection.

The option of vaccinating poultry against avian influenza (AI) as a control tool is gaining greater acceptance by governments and the poultry industry worldwide. One disadvantage about vaccination with killed whole-virus vaccines is the resulting inability to use common serologic diagnostic tests for surveillance to identify infected flocks. There has been considerable effort to develop a reliable test for the differentiation of infected from vaccinated animals (DIVA). The heterologous neuraminidase (NA) subtype DIVA approach has been used with some success in the field accompanied by an ad hoc serologic test. The traditional NA inhibition (NI) test can be used for all nine NA subtypes, but it is time consuming, and it is not designed to screen large numbers of samples. In this study, a quantitative NI test using MUN (2′-[4-methylumbelliferyl]-α-D-Nacetylneuraminic acid sodium salt hydrate) as an NA substrate was investigated as an alternative to the traditional fetuin-based NI test in a heterologous neuraminidase DIVA strategy. Serum NI activity was determined in chickens administered different vaccines containing different H5 and NA subtypes and challenged with a highly pathogenic avian influenza (HPAI) H5N2 virus. Prior to challenge, the NI DIVA test clearly discriminated between chickens receiving vaccines containing different antigens (e.g., N8 or N9) from control birds that had no NA antibody. Some birds began to seroconvert 1 wk postchallenge, and 100% of the vaccinated birds had significant levels of N2 NI activity. This activity did not interfere with the presence of vaccine-induced NI activity against N8 or N9 subtypes. The level of N2-specific NI activity continued to increase to the last sampling date, 4 wk postchallenge, indicating the potential use for the heterologous NA-based DIVA strategy in the field.

Vaccination against avian influenza (AI) virus, a powerful tool for control of the disease, may result in issues related to surveillance programs and international trade of poultry and poultry products. The use of AI vaccination in poultry would have greater worldwide acceptance if a reliable test were available that clearly discriminated between naturally infected and vaccinated-only animals (DIVA). Because the nonstructural protein (NS1) is expressed in influenza virus–infected cells, and it is not packaged in the virion, it is an attractive candidate for a DIVA differential diagnostic test. The aim of this work was to determine the onset of the antibody response to the NS1 protein in chickens infected with low pathogenic avian influenza (LPAI) virus, and to evaluate the diagnostic potential of a baculovirus-expressed purified NS1 protein in an indirect ELISA-based DIVA strategy. An antibody response against NS1 was first detected 3 wk after infection, but the antibody levels were decreasing rapidly by 5 wk after infection. However, most chickens did not have detectable antibodies in spite of high hemagglutination inhibition (HI) antibody titers in one group. In birds vaccinated with inactivated oil-emulsion vaccines, antibodies against NS1 were not detected before virulent challenge, and only a small percentage of birds seroconverted after homologous LPAI virus challenge. Vaccinated birds challenged with highly pathogenic AI showed a higher NS1 antibody response, but at most only 40% of birds seroconverted against NS1 protein by 3 wk after challenge. Because of the variability of seroconversion and the duration of the antibody response in chickens, the NS1 protein DIVA strategy did not perform as well as expected, and if this strategy were to be used, it would require sampling a higher number of birds to compensate for the lower seroconversion rate.

An inactivated H5N1 avian influenza (AI) vaccine generated by reverse genetics and containing the hemagglutinin and neuraminidase genes of the H5N1 A/goose/Guangdong/1/96 (GS/GD/06) virus has been used in domestic poultry in China and Vietnam as an important control strategy for H5N1 AI. The efficacy of this vaccine against early H5N1 isolates has been fully evaluated in chicken, ducks, and geese. However, there are no reports about its efficacy against H5N1 viruses recently isolated in China and Vietnam. In this study, groups of 3-wk-old specific-pathogen-free chickens were intramuscularly injected with one dose of the vaccine. Three weeks postvaccination, the chickens were intranasally challenged with 10⁵EID₅₀ of six different lethal H5N1 AI viruses: A/bar-headed goose/Qinghai/3/05 (clade 2.2), A/chicken/Shanxi/2/06 (CK/SX/06; clade 7), A/duck/Fujian/31/07 (clade 2.3.4), A/MDK/VN-HD/46/07 (clade 2.3.4), A/MDK/VN-CM/1185/06 (clade 1), and A/MDK/VN-CM/1159/06 (clade 1). Four out of 20 chickens challenged with the CK/SX/06 shed virus on day 5 and died on day 8 to 9 postchallenge. Chickens challenged with the remaining five viruses were completely protected (no disease signs, virus shedding, or deaths). These results indicate that the GS/GD/06-based vaccine provides sound protection against clade 1, 2.2, and 2.3.4 viruses, but not against the CK/SX/06 virus, which emerged in northern China in 2006.

Highly pathogenic avian influenza viruses (HPAIV) have historically caused disastrous damage to the poultry industry, and recently they have shown their zoonotic potential by causing human infections and deaths. Control and prevention of HPAIV are therefore important issues for both veterinary and human public health. In this study, we constructed a plasmid, pCAGGoptiH7, encoding a codon-optimized HA gene of the H7N1 avian influenza virus A/FPV/Rostock/34 (RK/34). To evaluate the vaccine efficacy of pCAGGoptiH7, groups of specific-pathogen-free (SPF) chickens were intramuscularly inoculated with one or two doses of 100 µg, 50 µg, or 10 µg of the plasmid in 3-wk intervals. Four weeks after the single vaccination or 2 wk after the second dose, all chickens were challenged with 100CLD₅₀ (chicken lethal dose) of highly pathogenic RK/34. After the single dose vaccination, only 90% of chickens were protected in all of the pCAGGoptiH7-immunized groups, although all of the chickens immunized generated detectable HI antibodies. After the second dose of vaccination, HI antibodies increased sharply, and chickens in the 100-µg and 50-µg pCAGGoptiH7-immunized groups were completely protected from virus challenge (no disease signs, no virus shedding, and no deaths). Low titers of virus shedding were detected in two out of ten chickens inoculated with two doses of 10-µg pCAGGoptiH7, although no disease or death was observed. These results provide a strong argument for the continued evaluation of this vaccine in field trials.

Infection with H9 avian influenza virus (AIV) and Newcastle disease virus (NDV) are two important causes of egg drop in layer and breeder poultry, leading to severe economic loss in the industry. Currently in China, inactivated H9 AIV vaccine and live attenuated NDV vaccine have to be repeatedly administered to prevent egg drop in layer animals. Using reverse genetics, we constructed a recombinant NDV expressing an H9 AIV hemagglutinin (HA) from an H9N2 field isolate, A/Chicken/Shandong/2/2007. The HA gene was inserted into the intergenic region between the phosphoprotein (P) and matrix (M) genes of the LaSota NDV vaccine strain. The recombinant virus stably expressing the HA gene, rL-H9, was found to be innocuous after intracerebral inoculation of 1-day-old chickens. A single dose of 10⁶ 50% egg infectious dose of the recombinant virus intranasally inoculated into chickens induced high levels of NDV- and AIV H9-specific hemagglutination-inhibition antibody. Complete protection from clinical disease and mortality against challenge with a lethal dose of velogenic NDV was observed in chickens and 90% of chickens were protected from clinical disease, mortality, and virus shedding against challenge with homologous H9N2 AIV. Our results suggest that recombinant NDV is suitable as a potential bivalent live attenuated vaccine against both NDV and H9 AIV infection in poultry.

Live attenuated vaccines can mimic natural infection and induce humoral and cellular immune response. However, the possibility of reassortment between vaccine viruses and field isolates and of mutations from low-pathogenic to highly pathogenic viruses has prevented the use of live attenuated strains as poultry vaccines. In ovo vaccination using live attenuated strains that can undergo limited replication cycles would be a better option, because these strains can be used for mass vaccination without spreading or reassorting with other viruses. Our previous study demonstrated that two influenza nonstructural (NS) variant viruses are highly attenuated and immunogenic in chickens, making them potential live vaccine candidates. In this study, we tested whether NS variants could be used as in ovo vaccines alone or in combination with temperature-sensitive (ts) mutations. In addition, we also tested the effect of different hemagglutinin (HA) subtypes on in ovo vaccination of NS variants. Our results demonstrated that NS variants alone or in combination with ts mutations were not attenuated enough to be used for in ovo vaccination. We also observed variable effects of different HA subtypes in the same NS deletion variant backbone on hatchability. However, even with substitution of HA subtypes, NS variant–inoculated eggs still had lower hatchability compared to the mock control group, indicating that the high virulence of NS variant backbone strain in eggs might have affected the results.

Highly pathogenic avian influenza viruses (AIVs) are Select Agents in the United States and are required to be handled in bio-containment level-3 enhanced (BSL3 ) facilities. Using a reverse genetics system, we attenuated a highly pathogenic virus, with the goal of making it low pathogenic and having it delisted as a Select Agent so that it could be handled in a bio-containment level-2 facility for diagnostic or vaccine production applications. We utilized two approaches to attenuate the target AIV by mutating the highly pathogenic hemagglutinin (HA) cleavage site to be low pathogenic and by replacing the full-length NS gene segment with a naturally truncated 124–amino acid NS1 coding gene from A/turkey/Oregon/73 (H7N3) virus (tkOR71 trNS1). To delist an AIV so that it can be handled in a BSL2 facility, the amino acid sequence of the HA cleavage site of the rescued virus must be confirmed to be compatible with a low-pathogenic AIV; it should not plaque in cell culture without supplementation of exogenous trypsin; and intravenous pathotyping in 4–6-wk-old specific-pathogen-free chickens must confirm that the virus is low pathogenic. The candidate A/duck/Vietnam/Baclieu/09/07 (rH5N1/PR8/trNS1) virus with five PR8 internal genes, tkOR71 trNS1 gene, and A/chicken/Indonesia/7/03 N1 neuraminidase gene was constructed. The virus was shown to not plaque in cell culture without addition of trypsin. The virus was low pathogenic in the standard intravenous pathotyping test (IVPI  =  0) and also caused no disease in a separate intranasal inoculation test in 4-wk-old specific-pathogen-free chickens, thus demonstrating that the virus is suitable for deselection.

High pathogenicity avian influenza H5N1 has become an endemic poultry disease in several Asian countries, including Vietnam. Recently, clade 7 H5N1 viruses of the Eurasian lineage were isolated from chickens seized at ports of entry in Lang Son Province, Vietnam. Extensive nucleotide and amino acid divergence across the hemagglutinin (HA) protein gene of these isolates in comparison to previously described clade 7 viruses was identified. Clade 7 viruses are antigenically distinct from contemporary strains of H5N1 known to circulate in Vietnamese poultry (clade 1 and clade 2.3.4). Subsequent surveillance of sick poultry in live poultry markets in Hai Duong Province identified additional clade 7 isolates with HA genes very similar to the group B virus cluster detected previously at the Lang Son Province border. Antigenic analysis of the isolates from the live bird markets revealed significant cross-reactivity only between those clade 7 viruses belonging to the same subgroups. To meet pandemic response preparedness objectives, we have developed a reassortant virus from A/chicken/Vietnam/NCVD-016/2008, which could be used as a new prepandemic vaccine candidate for veterinary or human vaccination, should the need arise. Findings from these studies indicate that viruses with clade 7 HA have continued to evolve in Southeast Asian poultry, leading to significant antigenic drift relative to other H5N1 viruses currently circulating in Vietnam.

Influenza A strains emerging from wild birds are a constant threat to South Africa's valuable ostrich industry. In 2004 and again in 2006, low pathogenicity avian influenza H5N2 strains introduced from a wild bird reservoir mutated in ostriches to high pathogenicity avian influenza (HPAI), with serious economic consequences and export bans imposed by the European Union. Although no outbreaks of notifiable avian influenza have occurred in South Africa since 2006, the H9N2 virus caused a localized outbreak where ostriches displayed symptoms of green urine, depression, and mild morbidity. Most recently, an outbreak of H10N7 in farmed Pekin ducks (Anas platyrhynchos domestica) caused increased mortalities, but this was exacerbated by a secondary Escherichia coli infection, because an intravenous pathogenicity index of 0.00 was recorded. Each of the eight gene segments of the five strains isolated from 2007 to 2009 from farmed ostriches in the Oudtshoorn region (H6N8, H9N2), Pekin ducks (H10N7, Joostenburgvlakte region), and wild Egyptian geese (Alopochen aegypticus; H1N8, Baberspan wetlands; H4N2, Oudtshoorn region) were sequenced, genetically analyzed, and compared to previous South African isolates and viruses in the public data banks. An H5N8 strain was also detected by reverse-transcription PCR in cloacal swabs from swift terns (Sterna bergii) in the Mosselbaai region during 2007, although a virus could not be isolated. Initial phylogenetic results indicate that H6N8 and H9N2 ostrich and H10N7 Pekin duck viruses originated in the wild bird population that is geographically dispersed throughout southern Africa, based on the reassortment of viral genes from birds sampled outside of the ostrich farming areas. No evidence of internal genes associated with Asian HPAI H5N1 strains were detected in the South African isolates.

Since 1999, the Italian poultry production system has experienced several outbreaks of avian influenza (AI), mainly located in northeastern Italy. This paper describes the low pathogenicity (LP) AI outbreaks detected during the surveillance activities implemented in 2007–08. From May to October 2007, ten rural and hobby poultry farms were infected by an LPAI virus of the H7N3 subtype. In August–October 2007, the H7N3 LPAI virus was introduced into the industrial poultry sector with the involvement of six meat turkey farms. Phylogenetic analysis of the hemagglutinin gene indicated that all but one of the H7N3 virus strains had a high level of homology (98.7%–99.8%). Furthermore, in August 2007, an LPAI H5N2 virus was identified in a free-range geese and duck breeder flock. The hemagglutinin and neuraminidase genes showed a high level of homology (99.8% and 99.9%, respectively) with H5N2 LPAI viruses isolated from mallards in July 2007 in the same area, suggesting a possible introduction from the wild reservoir. All the birds (in total 129,386) on the infected poultry farms were culled. The prompt implementation of AI control measures, including the enforcement of a targeted emergency vaccination plan, allowed the rapid eradication of infection. In 2008, three LPAI viruses (two H7N1 and one H5N1) were identified in dealer/rural farms. The surveillance activity implemented in this area allowed the prompt detection of LPAI viruses of the H5 and H7 subtypes in the rural sector, which, as observed in the 2007 epidemic, might be the source of infection for industrial poultry.

Highly pathogenic avian influenza A virus (H5N1) has diverged antigenically and genetically since its initial detection in Asia in 1997. Viruses belonging to clade 2.2 in particular have been reported in numerous countries with the majority occurring in Egypt. Previous reports identified antigenic similarities between viruses belonging to clade 2.2. However, poultry and human viruses isolated in northern Egypt during 2007 and 2008 were found to be antigenically distinct from other clade 2.2 viruses from this country. Genetic analysis of the hemagglutinin revealed a high degree of nucleotide and amino acid divergence. The antigenic changes in Egyptian viruses isolated during 2007–08 necessitated that two of these strains be considered as potential H5N1 pre-pandemic vaccine candidates.

This paper describes the results of the molecular and phylogenetic analysis of seven highly pathogenic avian influenza (HPAI) H5N1 strains isolated in 2006 (n  =  5) and 2007 (n  =  2) from wild birds and poultry in Poland. The whole genome sequence of these isolates was determined. All of the isolates possessed the hemagglutinin (HA) cleavage site sequence PQGERRRKKR*GLF typical of HPAI. Molecular markers associated with increased adaptation and virulence in mammals, as well as susceptibility to neuraminidase inhibitors, were revealed in the HA, neuraminidase (NA), and PB2 proteins. Based on the sequencing results related to the HA and NA genomic segments, H5N1 viruses circulating in Poland all belong to lineage 2.2. However, isolates isolated in 2006 were genetically distinct from those isolated in 2007 and grouped in different sublineages. H5N1 viruses isolated from wild birds in 2006 are almost identical to each other (99.9% HA; 99.6%–100% NA), and they are grouped within a cluster of viruses isolated in Germany from wild and domestic birds and mammals in 2006. Isolates from 2007 are also closely related to each other (nucleotide homologies 99.9% and 100% for HA and NA, respectively), and they are grouped together with isolates from wild and domestic birds collected in Eastern and Central Europe (Romania, Germany), and the Middle East (Kuwait, Saudi Arabia). Phylogenetic analysis of the sequences related to the internal proteins confirmed the results obtained for the HA and NA genes. Overall, the results indicate that HPAI H5N1 in Poland in 2006–07 was caused by at least two separate incursions of genetically distinct viruses.

A comprehensive avian influenza control program was established for the New York live bird market (LBM) system. Its purpose was to eliminate avian influenza virus (AIV) from the marketing system. The application of science-based surveillance, improved diagnostic performance, voluntary efforts of the LBM owners, and regulatory enforcement have resulted in the elimination of an H7 low pathogenic AIV (LPAIV) that had persisted in the LBM system for 13 yr. Although sporadic introductions of H5N2 LPAIVs have occurred, successful control measures have not allowed this virus to become established within the system.

The Eurasian-lineage H5N1 highly pathogenic avian influenza (HPAI) virus caused widespread outbreaks in Egypt in 2006 and eventually become enzootic in poultry. Although outbreaks have a seasonal pattern, with most occurring during the cooler winter months, it remains unclear whether this seasonality reflects virus maintenance within Egypt or yearly introductions of the virus into the country. To evaluate the epidemiology of H5N1 HPAI in Egypt, sequence analysis of the hemagglutinin (HA) and neuraminidase (NA) genes of selected Egyptian isolates from early 2006 to 2008 was conducted. The data from this study identifies distinct genetic markers in both HA and NA genes and suggests grouping Egyptian isolates into two major HA isolate sublineages from 2006 to 2008 and into three smaller, emergent subgroups. The NA phylogenetic and sequence analysis showed a similar pattern, except that two of the emergent groups from the HA phylogenetic tree clustered together, evidence of likely reassortment. The different subgroups did not appear to segregate by relation to the date of isolation, to the species of origin, nor to the geographic location of the viruses. The conclusion is that H5N1 is continuing to mutate with multiple heterogenic strains persisting in Egypt.

The first outbreak of H5N1 highly pathogenic avian influenza (HPAI) in the Kingdom of Saudi Arabia (KSA) occurred in two “backyard” flocks of Houbara bustards and falcons in February 2007. Subsequent outbreaks were seen through the end of 2007 in “backyard” birds including native chickens, ostriches, turkeys, ducks, and peacocks. From November 2007 through January 2008, H5N1 HPAI outbreaks occurred in 19 commercial poultry premises, including two broiler breeder farms, one layer breeder farm, one ostrich farm, and 15 commercial layer farms, with approximately 4.75 million birds affected. Laboratory diagnosis of all H5N1-positive cases was conducted at the Central Veterinary Diagnostic Laboratory (CVDL) in Riyadh, Saudi Arabia. A combination of diagnostic tests was used to confirm the laboratory diagnosis. A rapid antigen-capture test and real-time reverse transcriptase–PCR (rtRT-PCR) assay on clinical and field specimens were conducted initially. Meanwhile, virus isolation in specific-pathogen-free embryonating chicken eggs was performed and was followed by hemagglutinin (HA) and hemagglutination inhibition tests, then rapid antigen-capture and rtRT-PCR tests on HA-positive allantoic fluid samples. In most HPAI cases, a complete laboratory diagnosis was made within 24–48 hr at the CVDL. Saudi Arabian government officials made immediate decisions to depopulate all H5N1-affected and nonaffected flocks within a 5-km radius area and applied quarantine zones to prevent the virus from spreading to other areas. Other control measures, such as closure of live bird markets and intensive surveillance tests on all poultry species within quarantine zones, were in place during the outbreaks. As a result, the HPAI outbreaks were quickly controlled, and no positive cases were detected after January 29, 2008. The KSA was declared free of HPAI on April 30, 2008, by the World Animal Health Organization.

An outbreak of highly pathogenic avian influenza (HPAI) virus subtype H5N1 was first diagnosed in a “backyard” flock of peafowl (Pavo cristatus) raised on palace premises in the Kingdom of Saudi Arabia in December 3, 2007. The flock consisted of 40 peafowl, and their ages ranged from 3 to 5 years old. Affected birds suffered from depression, anorexia, and white diarrhea. Four dead birds were submitted for HPAI diagnosis at the Central Veterinary Diagnostic Laboratory in Riyadh. Brain and liver tissues and tracheal and cloacal swabs were taken from the dead birds and processed for a real-time reverse transcriptase (RT)-PCR test and virus isolation in specific-pathogen-free embryonating chicken eggs. The H5N1 subtype of avian influenza virus was isolated from the four dead birds and identified by a real-time RT-PCR before and after egg inoculation. The virus isolates were characterized as HPAI H5N1 virus by sequencing analysis. Phylogenetic comparisons revealed that the H5N1 viruses isolated from peafowl belong to the genetic clade 2.2 according to the World Health Organization nomenclature. The peafowl H5N1 virus falls into 2.2.2 sublineage II and clusters with the H5N1 viruses isolated from poultry in Saudi Arabia in 2007–08.

Newcastle disease virus (NDV) and avian influenza virus (AIV) are pathogens of major economic and social importance, and the diseases they cause are often devastating, particularly in domestic poultry. Both viruses are naturally found in a wide variety of wild birds, particularly aquatic species, where asymptomatic infection typically occurs. Wild birds are therefore considered to be a natural reservoir for both viruses. Wild birds kept in captivity are in an environment that promotes transmission of infection with both influenza and Newcastle disease viruses. This report describes a survey for the detection of antibodies against Newcastle disease and avian influenza A viruses using the hemagglutination inhibition test in samples from 88 wild birds from 38 species in four Bulgarian zoos. Samples with positive results against NDV were also tested against avian paramyxovirus type 3 (APMV-3). Real-time reverse-transcriptase PCR was also performed to detect viral RNA of NDV and AIV among 127 wild birds from 57 species from the same zoos. In 13 samples from seven avian species (ten birds from the family Phasianidae, two from the family Numidae, and one from the family Columbidae), antibodies against APMV-1 were detected. Seven birds, whose sera were APMV-1 positive, had been vaccinated. The other six birds (five Phasianidae representatives and one of the Columbidae family) had no immunization history. No antibodies against both H5 and H7 AIV and against APMV-3 were detected, and no RNA of NDV and AIV were detected.

A serologic survey for antibodies against H5 subtype influenza virus in 605 apparently healthy local chickens using a hemagglutination inhibition test was carried out in 12 local government areas of Kaduna state, Nigeria. An overall prevalence of 18.1% was recorded, with a higher prevalence of 27.3% in six local government areas that have not reported outbreaks of highly pathogenic avian influenza (HPAI) H5N1 virus and a lower prevalence of 7.5% in six local government areas that had reported and confirmed outbreaks of HPAI H5N1 virus between 2006 and 2007. There was association between the presence of ducks and detection of H5 antibodies (P  =  0.000, odds ratio  =  0.22). The implication of this finding is discussed, although a virologic investigation to verify the findings of this study is highly recommended.

The Caribbean region is considered to be at risk for avian influenza (AI) due to a large backyard poultry system, an important commercial poultry production system, the presence of migratory birds, and disparities in the surveillance systems. The Caribbean Animal Health Network (CaribVET) has developed tools to implement AI surveillance in the region with the goals to have 1) a regionally harmonized surveillance protocol and specific web pages for AI surveillance on  http://www.caribvet.net, and 2) an active and passive surveillance for AI in domestic and wild birds. A diagnostic network for the Caribbean, including technology transfer and AI virus molecular diagnostic capability in Guadeloupe (real-time reverse transcription-polymerase chain reaction for the AI virus matrix gene), was developed. Between 2006 and 2009, 627 samples from four Caribbean countries were tested for three circumstances: importation purposes, following a clinical suspicion of AI, or through an active survey of wild birds (mainly waders) during the southward and northward migration periods in Guadeloupe. None of the samples tested were positive, suggesting a limited role of these species in the AI virus ecology in the Caribbean. Following low pathogenic H5N2 outbreaks in the Dominican Republic in 2007, a questionnaire was developed to collect data for a risk analysis of AI spread in the region through fighting cocks. The infection pathway of the Martinique commercial poultry sector by AI, through introduction of infected cocks, was designed, and recommendations were provided to the Caribbean Veterinary Services to improve cock movement control and biosecurity measures. The CaribVET and its organization allowed interaction between diagnostic and surveillance tools on the one hand and epidemiologic studies on the other, both of them developed in congruence with regional strategies. Together, these CaribVET activities contribute to strengthening surveillance of avian influenza virus (AIV) in the Caribbean region and may allow the development of research studies on both AI risk analysis and on AIV ecology.

This article explores the economic and related institutional issues at macro and micro levels, in different production systems and in different countries that influence avian influenza (AI) management and control. It does this by examining three groups of stakeholders with different agendas and concerns. For the “international community,” the overriding driver has been and still is concern for human safety. This is reflected in the high level of contributions to emergency response programs, a strong focus on pandemic prevention and preparedness, and the pressure put on countries to develop prevention and control plans. For the most influential countries and companies in the global poultry sector, those that control the largest commercial poultry populations, trade growth and stability are major concerns. Private investment in biosecurity, reorganization of supply chains, and an increasing interest in compartments are all indications of a perceived need to secure the boundaries. Poor poultry-keeping households must focus on day-to-day livelihoods and food security, whereas small-scale commercial producers are driven by small margins and short credit cycles. Although these people operate a little differently, they have in common a necessity to focus on the short term and a limited willingness and ability to invest in their flocks. There is also very little information that we can provide either of them on financially viable ways to upgrade their enterprises. Noncompliance or partial compliance with AI regulations often makes good economic sense. Different highly pathogenic AI management and control measures are economically viable in different circumstances. The article discusses the positive and less-positive impacts created by each stakeholder perspective and the conflicts and trade-offs that can arise, and suggests some approaches for reconciling differences and thus improving AI control.

The World Organisation for Animal Health (OIE)/United Nations Food and Agriculture Organization (FAO) joint network of expertise on animal influenza (OFFLU) includes all ten OIE/FAO reference laboratories and collaborating centers for avian influenza, other diagnostic laboratories, research and academic institutions, and experts in the fields of virology, epidemiology, vaccinology, and molecular biology. OFFLU has made significant progress in improving its infrastructure, in identifying and addressing technical gaps, and in establishing associations among leading veterinary institutions. Interaction with the World Health Organization (WHO) Global Influenza Program is also critical, and mechanisms for permanent interaction are being developed. OFFLU played a key role in the WHO/OIE/FAO Joint Technical Consultation held in Verona (October 7–9, 2008), which provided an opportunity to highlight and share knowledge and identify potential gaps regarding issues at the human-animal interface for avian influenza. OFFLU experts also contributed to the working group for the Unified Nomenclature System for H5N1 influenza viruses based on hemagglutinin gene phylogeny (WHO/OIE/FAO, H5N1 Evolution Working Group, Towards a unified nomenclature system for highly pathogenic avian influenza virus (H5N1) in Emerging Infectious Diseases 14:e1, 2008). OFFLU technical activities, led by expert scientists from OIE/FAO reference institutions and coordinated by OIE and FAO focal points, have been prioritized to include commercial diagnostic kit evaluation, applied epidemiology, biosafety, vaccination, proficiency testing, development of standardized reference materials for sera and RNA, and issues at the human-animal interface. The progress to date and future plans for these groups will be presented. OFFLU is also involved in two national projects implemented by FAO in Indonesia and Egypt that seek to establish sustainable mechanisms for monitoring virus circulation, including viral characterization, and for streamlining the process to update poultry vaccines for avian influenza.

Global epidemics of H5N1 highly pathogenic avian influenza have highlighted aspects that are crucial when considering effective and rapid control of animal diseases, including zoonoses. Experience has shown that without strong and well-governed veterinary services, effective surveillance and control can be challenging at best. For an effective response, veterinary services need access to expertise, resources, and appropriate legislation. Within its mandate to “improve animal health worldwide” the World Organisation for Animal Health is providing recommendations and taking action to improve the strength and governance of veterinary services worldwide.

Since 2006, a collaborative group of egg industry, state, federal, and academia representatives have worked to enhance preparedness in highly pathogenic avian influenza (HPAI) planning. The collaborative group has created a draft egg product movement protocol, which calls for realistic, science-based contingency plans, biosecurity assessments, commodity risk assessments, and real-time reverse transcriptase–PCR testing to support the continuity of egg operations while also preventing and eradicating an HPAI outbreak. The work done by this group serves as an example of how industry, government, and academia can work together to achieve better preparedness in the event of an animal health emergency. In addition, in the event of an HPAI outbreak in domestic poultry, U.S. consumers will be assured that their egg products come from healthy chickens.

This research assessed the direct economic effects of the 2005–06 HPAI outbreaks on contracted turkey producers in Turkey in 2007. The data were obtained from 71 randomly selected, contracted turkey farms (producing 23% of the national turkey meat in Turkey in 2005) from five provinces for four integrated firms, which account for 67% of the national turkey production. Each farm was visited once for an oral interview conducted by the authors in 2007, using a questionnaire survey. The financial data before and after highly pathogenic avian influenza (HPAI) H5N1 outbreak periods were obtained from available financial records. Changes in production and economics parameters before and after the HPAI H5N1 outbreak periods were compared. In the analyses, the “before the HPAI H5N1 outbreak” period was stated as October 1, 2004, to May 31, 2005, whereas the “after the HPAI H5N1 outbreak” period was stated as the 8-mo between October 1, 2005, and May 31, 2006. The research revealed that changes in the technical parameters (number of hired labor, feed conversion rate, mortality rate, and the length of fattening period) were not found to be statistically significant at P > 0.05. However, there were severe effects of the HPAI H5N1 outbreaks on the economic parameters of the turkey production. The contracted turkey producers lost on average 0.9 cycles (38%) of production, and their management fees were reduced by 9.3% in the 8 mo after the outbreaks. As a result, the production level and enterprise income declined by 36% and 39%, respectively. About 93% of the producers did not do any other supplementary work during the idle production period; 59% of the producers had to use on average 4970TL (US$3200) from their personnel saving during the HPAI H5N1 outbreaks. About 62% of the producers stated that they had been considering expanding their businesses, but suspended the idea because of the outbreak, and 80% of the producers increased the biosecurity measures after the outbreaks. The futures of the contracted turkey producers are fully dependent on those of the integrated firms. Any negative effects on the latter appear to be directly transferred to the former. However, the government neglected contracted producers in the HPAI compensation programs.

Influenza surveillance in wild birds has established that the aquatic birds of the world are the source of influenza A viruses, which occasionally spread to domestic avian species and to mammals, including humans, and cause mild to severe disease. With the realization that the pandemics of influenza in poultry and people originate from inapparent infections of aquatic birds, including the highly pathogenic H5N1 virus, much more attention has been given to understanding the ecology of influenza in wild aquatic birds. This article deals with the major events establishing the role of wild birds in the natural history of influenza and with some of the unresolved issues. These include 1) whether all H5 and H7 influenza viruses have high pandemic potential, 2) whether avian influenza (AI) is exchanged between Eurasia and the Americas, and 3) whether the highly pathogenic H5N1 AI virus is now being perpetuated in wild birds, one of the most important unresolved issues. Continued surveillance of wild birds for influenza is essential to resolve the many unanswered questions concerning the zoonotic spread of influenza and pandemicity.

Surveillance of wild birds for avian influenza viruses has been compulsory in the European Union (EU) since 2005, primarily as a means of detecting H5N1 highly pathogenic avian influenza (HPAI) virus and of monitoring the circulation of low pathogenicity avian influenza (LPAI) virus H5 and H7 strains. In 2007, 79,392 wild birds were tested throughout the EU. H5N1 HPAI was detected in 329 birds from four Member States (MS); affected birds were almost entirely of the orders Podicipediformes (grebes) and Anseriformes (waterfowl) during the summer months. LPAI was detected in 1485 wild birds among 21 MS. A total of 1250 birds were positive for influenza A but were not discriminated any further; LPAI H5 was detected in 105 birds, exclusively of the order Anseriformes. LPAI H7 was detected in seven birds. LPAI of other subtypes was found in 123 birds. Epidemiologic evidence and phylogenetic analysis of H5N1 viruses indicate that H5N1 did not appear to persist in the EU from 2006 but was reintroduced, probably from the Middle East.

Due to concerns that high pathogenicity avian influenza would enter into the United States, an interagency strategic plan was developed to conduct surveillance in wild birds in order to address one of the possible pathways of entry. The USDA and state wildlife agencies participated in this effort by collecting samples from 145,055 wild birds from April 2006 through March 2008 in all 50 states. The majority (59%) of all wild bird samples was collected from dabbling ducks, and 91% of H5 detections using real-time reverse transcriptase polymerase chain reaction (rRT-PCR) were in dabbling ducks. Apparent prevalence of H5 by rRT-PCR in all birds sampled was 0.38%. Most (48%) H5 detections were found in mallards (Anas platyrhynchos). Thirty-three virus subtypes were identified; H5N2 was the most prevalent subtype and accounted for 40% of all virus isolations. We present the virus subtypes obtained from the national surveillance effort and compare them with research results published from various countries.

Avian influenza virus (AIV) was studied in ring-billed gulls (Larus delawarensis) in one breeding colony on Lake Erie in 2000, and two on Lake Ontario in both 2000 and 2004. Antibodies to H13 AIV were detected in 92% of adults in 2000 and 82% in 2004. Antibody prevalence in 3-wk-old chicks was 5%–30% (overall 15%) in 2000 and 21% and 76% (overall 48%) in 2004. In 5-wk-old chicks, antibody prevalence was 23%–75% (overall 53%) in 2000 and 53% and 79% (overall 66%) in 2004. Geometric mean antibody titers at 3 and 5 wk did not differ in 2000, but increased significantly at one colony in 2004. In 2000, overall prevalence of AIV isolation from cloaca in embryonated chicken eggs was 32% (3 wk old), 13% (5 wk old), and 0 (adults), but AIV was also isolated from kidney and lung in a high proportion of tissues cultured from 3-wk-old birds in one colony. Isolates from cloaca were characterized as subtype H13 by serology; all 15 tested for neuraminidase were H13N6. However, three AIV detections considered on the basis of nucleotide sequence to be subtype H16 were among the 28 detected retrospectively by PCR in archived cloacal swabs; the remainder were subtype H13. Outcome of virus isolation was not related to presence of antibody titers in chicks. The presence of antibody to AIV in chicks was associated significantly with inflammation in heart, kidney, pancreas, and liver. AIV was not isolated in 2004. AIV infected chicks annually within the first 3 wk of life, ultimately infecting the majority of birds in most colonies, but did not appear to cause clinical disease.

Denmark forms a geographical bottleneck along the migration route of many water birds breeding from northeastern Canada to north Siberia that gather to winter in Europe and Africa. Potentially, the concentration of such large numbers of water birds enhances the risk of avian influenza virus (AIV) introduction to domestic poultry. In 2003, Denmark initiated a nationwide survey of AIV in wild birds and mallards reared for shooting. Partial sequence analysis of the six internal genes from a total of 12 low pathogenic (LP) AIV isolates obtained in 2003 showed that genes from these viruses were closely related with genes from AIV circulating in northern Europe. For the Danish sequences only the PB2 and NS genes differ, so they cluster to more than one cluster in the phylogenetic trees. In spring 2006, highly pathogenic (HP) AIV H5N1 was detected in 44 cases of wild birds in Denmark. Sequence analysis of the HP H5N1 virus genome showed that it was not related to the LPAIV isolated previously, but closely related to the HPAIV H5 (Asian type) detected in the rest of Europe at that time. Even though only partial sequences were applied, this gave the idea for future full-length sequence studies.

We examined whether host traits influenced the occurrence of avian influenza virus (AIV) in Anatidae (ducks, geese, swans) at wintering sites in California's Central Valley. In total, 3487 individuals were sampled at Sacramento National Wildlife Refuge and Conaway Ranch Duck Club during the hunting season of 2007–08. Of the 19 Anatidae species sampled, prevalence was highest in the northern shoveler (5.09%), followed by the ring-necked duck (2.63%), American wigeon (2.57%), bufflehead (2.50%), greater white-fronted goose (2.44%), and cinnamon teal (1.72%). Among host traits, density of lamellae (filtering plates) of dabbling ducks was significantly associated with AIV prevalence and the number of subtypes shed by the host, suggesting that feeding methods may influence exposure to viral particles.

Infection with highly pathogenic avian influenza virus H5N1 occurred for the first time in Denmark in 2006 during the last part of the European epidemic that mainly affected migrating wild birds. The total number of Danish wild bird cases was 45, of which only one was found through active surveillance using fecal sampling from resting areas for migrating species, whereas passive surveillance of dead wild birds provided 44 cases. One backyard, mixed poultry flock also became infected late in the epidemic. This study describes the spatial and temporal distribution of cases, initially characterized by a spatial-temporal cluster of affected tufted ducks that led to further spread to other wild bird species in the vicinity. The surveillance data also indicate an apparent die-off of the regional epidemic. As a tool in visualizing the spatial and temporal development of the epidemic, a prototype avian influenza (AI) BioPortal was used to provide online web-based access to the data. The AI BioPortal tools include mapping, graphing, phylogenetic tree construction, playback scenarios, and visualization of results of temporal-spatial analyses. Several of the features of this surveillance system compare favorably to the design of existing national and international surveillance information systems, and the AI BioPortal may become a useful tool for disease surveillance and for decision support in the event of future AI epidemics, both at national and international levels.

A multi-agency, Canada-wide survey of influenza A viruses circulating in wild birds, coordinated by the Canadian Cooperative Wildlife Health Centre, was begun in the summer of 2005. Cloacal swab specimens collected from young-of-year ducks were screened for the presence of influenza A nucleic acids by quantitative, real-time reverse transcription-polymerase chain reaction (RRT-PCR). Specimens that produced positive results underwent further testing for H5 and H7 gene sequences and virus isolation. In addition to live bird sampling, dead bird surveillance based on RRT-PCR was also carried out in 2006 and 2007. The prevalence of influenza A viruses varied depending on species, region of the country, and the year of sampling, but generally ranged from 20% to 50%. All HA subtypes, with the exception of H14 and H15, and all NA subtypes were identified. The three most common HA subtypes were H3, H4, and H5, while N2, N6, and N8 were the three most common NA subtypes. H4N6, H3N2, and H3N8 were the three most common HA–NA combinations. The prevalence of H5 and H7 subtype viruses appears to have a cyclical nature.

Situated at the crossroads of numerous migratory routes of Palaearctic birds, the Camargue is considered a high-risk area for the introduction and transmission of numerous avian-borne pathogens. We investigated the epidemiologic cycles of avian influenza viruses (AIVs) in the local bird community by performing regular sampling on a large variety of bird species during 11 consecutive months in 2006–07. To detect the presence of AIV, SYBR green reverse transcriptase–PCR targeting the M gene was performed on 2901 samples from 66 bird species. A clear seasonal pattern of AIV circulation in ducks was observed during autumn and winter, with higher prevalence rates in early fall. Our results also support an absence of circulation of AIV in passerine birds during spring and the wintering periods. Finally, even if the prevalence of infection was very low, AIVs were found in gulls in breeding colonies, indicating a possible specific circulation in spring in these birds.

Bulgaria has a unique geographic position in Europe, with two migratory wild bird routes, Via Pontica and Via Aristotelis, passing through the country. Via Pontica is the second-largest migration route in Europe, with hundreds of thousands of birds, representing more than 110 species, wintering in lakes by the Black Sea and the wetlands near the Danube River. Via Aristotelis is situated in West Bulgaria along the Strouma and Mesta river valleys, and it is of regional importance for the Balkan Peninsula. In this study, we examined more than 2000 samples from wild birds from the orders Anseriformes, Ciconiiformes, Gruiformes, and Charadriformes in the period 2006–2008. We isolated three influenza viruses, subtypes H4N6, H7N7, and H10N7, all from mallards, Anas platyrhynchos. The H7N7 was isolated from a hunter-killed mallard at the river bank of Kamchia (Via Pontica). The cleavage site sequence of the hemagglutinin gene in the H7N7 isolate was PEIPKGR*GLF, which is characteristic of a low-pathogenic virus. The H4N6 isolates belonged to a mallard wintering along the Maritza River (Via Aristotelis). We detected the H10N7 virus in samples from mallards that inhabit the Ogosta River, one of the feeders of the Danube (Via Pontica). All these viruses were detected during the active migration of the birds, February–March.

The H7 subtype of avian influenza (AI) has the capability to evolve into a highly pathogenic AI (HPAI) virus. In this study, we have characterized the hemagglutinin (HA) genes of three avian H7N7 influenza A viruses isolated from healthy migratory mallards in Northern Europe in three different years to study the natural variation of these viruses in the natural reservoir. Phylogenetic analysis demonstrated that the H7 HA genes were all closely related to recent H7 isolates responsible for influenza outbreaks in poultry in Europe. The A/mallard/Sweden/S90735/2003 isolate clustered together with the HA gene of A/mallard/Netherlands/12/00/H7N3 (AY338460), which has been shown to be closely related to the H7N7 responsible for HPAI outbreaks in the Netherlands and Germany in 2003. In contrast, the HA gene of the two mallard strains A/mallard/Sweden/S90597/2005 and A/mallard/Sweden/100993/2008 were more related to the chicken strain isolated in domestic poultry in England in 2006, A/Ch/England/4054/2006/H7N3 (EF467826), and 2008, A/Ch/England/2008/H7N7 (214011964). Analysis of the deduced HA amino acid sequence shows two different HA cleavage sites in these isolates. Although these HA cleavage sites are consistent with a low pathogenicity AI, the cleavage sites appear to posses an increasing numbers of basic amino acids over time (PEIPKGRGLF in 2003 and 2005 and PEIPKKRGLF in 2008). The conclusion from this study is that H7 subtypes isolated from healthy mallards are closely related to the H7 subtypes that have caused recent influenza outbreaks in poultry in Europe.

Avian influenza is endemic in some species of wild birds and is generally believed to cause only an asymptomatic infection. These viruses are routinely transmitted from this wild bird reservoir to poultry in many areas all over the world. Low pathogenic avian influenza (LPAI) was previously reported in Egypt from different types of wild birds. This report describes the isolation and genetic characterization of H7N7 LPAI virus from a black kite (Milvus migrans), the first reported from this species, during surveillance done on wild birds in 2005. The black kite is a migratory bird that has breeding habitat in Europe and migrates in the winter to North Africa and the Middle East. Eight samples were collected in South Sinai, Egypt, and tested by virus isolation in embryonating chicken eggs. One sample had positive hemagglutination activity after the second passage in specific-pathogen-free embryos. Virus identification and characterization were done and the isolate was confirmed as H7N7 LPAI. The sequence data showed that this isolate was most closely related to European H7 strains isolated from domestic and wild birds.

Although aquatic habitats utilized by wild and domestic birds potentially can provide a bridge for avian influenza virus (AIV) transmission among many diverse hosts, the factors controlling environmental persistence and transmission via these habitats are poorly understood. AIV has been detected in water samples collected in the field, and under experimental laboratory conditions, these viruses can remain infective in water for periods of time that would be consistent with an environmental reservoir. However, the application of laboratory results to field realities is complicated by the complexity and scale of these systems. In this brief review, we present a summary of existing research on the environmental tenacity of AIV, provide an example of the challenges associated with the application of laboratory results to the field realities associated with detection of AIV from environmental sources, and identify gaps in our current understanding of the factors potentially affecting AIV infectivity in the environment, specifically from aquatic habitats utilized by wild birds.

Poyang Lake is situated within the East Asian Flyway, a migratory corridor for waterfowl that also encompasses Guangdong Province, China, the epicenter of highly pathogenic avian influenza (HPAI) H5N1. The lake is the largest freshwater body in China and a significant congregation site for waterfowl; however, surrounding rice fields and poultry grazing have created an overlap with wild waterbirds, a situation conducive to avian influenza transmission. Reports of HPAI H5N1 in healthy wild ducks at Poyang Lake have raised concerns about the potential of resilient free-ranging birds to disseminate the virus. Yet the role wild ducks play in connecting regions of HPAI H5N1 outbreak in Asia is hindered by a lack of information about their migratory ecology. During 2007–08 we marked wild ducks at Poyang Lake with satellite transmitters to examine the location and timing of spring migration and identify any spatiotemporal relationship with HPAI H5N1 outbreaks. Species included the Eurasian wigeon (Anas penelope), northern pintail (Anas acuta), common teal (Anas crecca), falcated teal (Anas falcata), Baikal teal (Anas formosa), mallard (Anas platyrhynchos), garganey (Anas querquedula), and Chinese spotbill (Anas poecilohyncha). These wild ducks (excluding the resident mallard and Chinese spotbill ducks) followed the East Asian Flyway along the coast to breeding areas in northern China, eastern Mongolia, and eastern Russia. None migrated west toward Qinghai Lake (site of the largest wild bird epizootic), thus failing to demonstrate any migratory connection to the Central Asian Flyway. A newly developed Brownian bridge spatial analysis indicated that HPAI H5N1 outbreaks reported in the flyway were related to latitude and poultry density but not to the core migration corridor or to wetland habitats. Also, we found a temporal mismatch between timing of outbreaks and wild duck movements. These analyses depend on complete or representative reporting of outbreaks, but by documenting movements of wild waterfowl, we present ecological knowledge that better informs epidemiological investigations seeking to explain and predict the spread of avian influenza viruses.

Recent literature has underestimated the number and taxonomic diversity of wild birds moving between Asia and North America. Our analyses of the major avian influenza (AI) host groups show that fully 33 species of waterfowl (Anatidae), 46 species of shorebirds (Charadriidae and Scolopacidae), and 15 species of gulls and terns (Laridae) are involved in movements from Asia to Alaska across northern oceans (Table 1). Our data suggest that about 1.5–2.9 million individuals in these important host groups move from Asia to Alaska annually. Among all of the host groups we consider most relevant for AI virus movement models in this region (waterfowl, shorebirds, and gulls and terns), it seems likely that thousands of AI-infectious birds may be involved in annual Asia-to-America migrations. Importantly, host availability in Alaska once these vectors arrive is also very high, representing at least 5–10 times more birds and infectious birds than the host populations moving from Asia to North America. Incorporating our data into a recent model of the global spread of the highly pathogenic H5N1 suggests that wild birds are a more likely source of this strain being brought into the United States than trade in domestic birds, although the latter remain a numerically more probable source of introduction into the New World. Our results should help in defining the key taxonomic, geographic, and seasonal factors involved in this complex intercontinental association of wild bird AI hosts. The next steps are to determine infection rates of low pathogenicity and highly pathogenic viruses among these hosts and to incorporate these into dynamic models.

Fifty-four strains of H5N1 highly pathogenic avian influenza (HPAI) virus were isolated from wild birds in the ecosystems of northern Eurasia and from poultry in the south of western Siberia (July 2005), at the mouth of Volga River (November 2005), at Uvs-Nur Lake on the boundary of the Great Lakes Depression in western Mongolia and the Tyva Republic of Russia (June 2006), in the vicinity of Moscow (February 2007), in the southeastern part of the Russian Plain (September 2007 and December 2007), and in the far east (April 2008) of the Russian Federation and were phenotypically characterized and deposited into the Russian state collection of viruses. Complete genome nucleotide sequences for 24 strains were obtained and deposited into GenBank. In all cases when strains were isolated from both wild birds and poultry in the same outbreak these strains were genetically closely related to each other. Until 2008 all HPAI H5N1 strains isolated in northern Eurasia clustered genetically with the viruses from Kukunor Lake (Qinghai Province, China), known as genotype 2.2 or the “Qinghai-Siberian” genotype. The viruses from the Qinghai-Siberian genotype have continued to evolve from those initially introduced into western Siberia in 2005 into two genetic groups: “Iran–North Caucasian” and “Tyva-Siberian.” In vitro replication potential (50% tissue-culture infectious dose in porcine embryo kidney) of Qinghai-Siberian strains decreased over time, which could reflect decreasing virulence. Comparison of genome sequences with biological characteristics of the respective strains permitted us to identify point mutations in PB2, PB1, PA, HA, NP, NA, M2, NS1, and NS2 that possibly influenced the level of replication potential. The HPAI H5N1 virus, which penetrated into the south of the Russian Far East in spring 2008, belonged to genotype 2.3.2.

In order to determine the actual prevalence of avian influenza viruses (AIVs) in wild birds in Bosnia and Herzegovina, extensive surveillance was carried out between October 2005 and April 2006. A total of 394 samples representing 41 bird species were examined for the presence of influenza A virus using virus isolation in embryonated chicken eggs, PCR, and nucleotide sequencing. AIV subtype H5N1 was detected in two mute swans (Cygnus olor). The isolates were determined to be highly pathogenic avian influenza (HPAI) virus and the hemagglutinin sequence was closely similar to A/Cygnus olor/Astrakhan/Ast05-2-10/2005 (H5N1). This is the first report of HPAI subtype H5N1 in Bosnia and Herzegovina.

Highly pathogenic (HP) avian influenza A viruses (AIVs) subtype H5N1 (subclade 2.2) were detected in wild birds during outbreaks in France during winter 2006 and summer 2007 in la Dombes wetlands (eastern France) and in Moselle wetlands (northeastern France), respectively. Blood samples from apparently healthy wild birds were collected in 2006 and 2007 from the end of the outbreak to several weeks after the influenza A outbreak inside and outside the contaminated areas, and in 2008 outside the contaminated areas. The samples were tested for the presence and/or quantitation of serum antibodies to influenza A subtypes H5 and N1 using hemagglutination inhibition tests (HITs), a commercial N1-specific enzyme-linked immunosorbent assay kit, and virus neutralization assay. In the HIT, low pathogenicity (LP) and HP H5 AIVs (belonging to H5N1, H5N2, and H5N3 subtypes) were used as antigens. One hundred mute swans were bled in the la Dombes outbreak area in 2006. During 2007, 46 mallards, 69 common pochards, and 59 mute swans were sampled in the Moselle outbreak area. For comparison, blood samples were also collected in 2007 from 60 mute swans from the Marne department where no HP H5N1 influenza A cases have been reported, and in 2008 from 111 sacred ibises in western France where no HP H5N1 influenza A infections in wild birds have been reported either. Mute swans (irrespective of their origin and time of sampling) and sacred ibises (from an area with no known outbreaks) had the highest prevalence of positive sera in the H5 HIT (49–69% and 64%, respectively). The prevalence of anti-H5 antibodies in mallards and common pochards was lower (28% and 27%, respectively). Positive H5- and N1-antibody responses were also significantly associated in swans (irrespective of their origin and time of sampling) and in sacred ibises. However, in swans from the area without outbreaks, the HIT titer against an H5N1 LPAIV was significantly higher than against an H5N1 2.2.1 HPAIV, whereas no difference could be shown for swans from the outbreak areas sampled in 2006 and 2007. These results suggest that ibises and swans from areas without declared outbreaks had acquired humoral immunity after AIV infections with subtypes H5 and N1 but independently from HP H5N1 infection. However, for swans living in outbreak areas, it cannot be excluded that this immunity might result from either a subclinical or a nonlethal infection by HP H5N1.

In April 2008 an avian influenza outbreak was diagnosed in Primorsky Krai, Russia, during the spring migration of wild birds, and A/Chicken/Primorsky/85/08 H5N1 isolate was recovered. The virus had more than 99% genetic identity with A/Whooper Swan/Hokkaido/1/08 H5N1 and A/Whooper Swan/Hokkaido/2/08 H5N1 viruses that were isolated in April 2008 in Japan. The amino acid sequence of the hemagglutinin cleavage site (PQRERRRKRGLF) and intravenous pathotyping index value (IVPI 2.80) were determined; on this basis the virus was characterized as highly pathogenic. The hemagglutinin gene of the virus was shown to belong to clade 2.3.2 while other genes (PB1, PB2, PA, NP, NA, M, NS) were characteristic of Fujian-like sublineage, recovered in the territory of Russia for the first time.

The influenza A/Mallard/Pennsylvania/10218/1984 (H5N2) virus is unable to replicate in 3-wk-old immunocompetent specific-pathogen-free chickens when a dose of 5 × 10⁶ 50% egg infectious dose/ml is used. In contrast, this mallard virus shows limited replication in 3-wk-old chickens that had been previously infected at 2 days of age with, and recovered from, the immunosuppressive agent infectious bursal disease virus (IBDV; herein referred to as IBDV chickens). This limited replication in IBDV chickens allowed for the serial passage of the mallard influenza virus in chickens. After 22 passages (P22) in IBDV chickens, the resulting chicken-adapted influenza virus replicated in both immunocompetent and IBDV chickens more efficiently than the mallard influenza virus. Analysis of the outcomes of infection and the lesions caused by the two viruses at the microscopic level in a time-point study showed that the P22 virus is more virulent than the parental mallard virus in both immunocompetent and IBDV chickens. Our studies provide evidence that a previous history of IBDV infection in chickens may render them more susceptible to avian influenza virus (AIV) infections, allowing for the potential introduction of AIVs in an otherwise resistant population.

Several previous reports and our studies show that waterfowl-origin influenza viruses can be more easily transmitted to domestic turkeys than chickens. Similarly, studies indicate turkeys to be better hosts for low pathogenic avian influenza viruses isolated from commercial poultry operations and live bird markets in comparison to chickens. Low 50% infectious-dose titers of wild bird as well as poultry-adapted viruses for turkeys further suggest that turkeys can be easily infected following a low-dose exposure. Also, interspecies transmission of swine influenza viruses to turkeys occurs frequently. These findings suggest the role of turkeys as suitable intermediate hosts that can be easily infected with influenza viruses of different origins and that turkeys can act as source of infection for other land-based poultry or even mammals.

The NS1 protein of influenza A viruses is known as a nonessential virulence factor inhibiting type I interferon (IFN) production in mammals and in chicken cells. Whether NS1 inhibits the induction of type I IFNs in duck cells is currently unknown. In order to investigate this issue, we used reverse genetics to generate a virus expressing a truncated NS1 protein. Using the low pathogenic avian influenza virus A/turkey/Italy/977/1999 (H7N1) as a backbone, we were able to rescue a virus expressing a truncated NS1 protein of 99 amino acids in length. The truncated virus replicated poorly in duck embryonic fibroblasts, but reached high titers in the mammalian IFN-deficient Vero cell line. Using a gene reporter system to measure duck type I IFN production, we showed that the truncated virus is a potent inducer of type I IFN in cell culture. These results show that the NS1 protein functions to prevent the induction of IFN in duck cells and underline the need for a functional NS1 protein in order for the virus to express its full virulence.

Until 2002, H5N1 highly pathogenic avian influenza (HPAI) viruses caused only mild respiratory infections in ducks. Since then, new viruses have emerged that cause systemic disease and high mortality in ducks and other waterfowl. Studies on HPAI virus pathogenicity in ducks have been limited, and there is no clear explanation of why the pathogenicity of some H5N1 HPAI viruses has increased. The nonstructural protein 1 (NS1 protein) is known to suppress immune responses in influenza virus–infected hosts affecting virus pathogenesis. In order to determine if the NS1 protein contributes to increased virulence in ducks, single-gene reassortant viruses were generated. Exchanging the NS genes from A/Ck/HK/220/97 (a virus that produces mild disease in ducks) and A/Dk/VN/201/05 (a very virulent virus for ducks) in the rEgret/02 background (a recombinant virus derived from A/Egret/HK/757.2/02, a highly pathogenic virus in ducks) resulted in decreased mean death times compared to infection with the rEgret/02 virus in ducks, but the change was not statistically significant. Infection with the reassortant viruses affected the expression of immune-related genes in spleens and lungs when compared to controls, but when compared among them, the expression of the duck genes was similar. Furthermore, virus titers in spleen, lung, and brain as well as antigen distribution in various tissues were similar in ducks infected with the reassortant viruses. All together these data show that, under these experimental conditions, exchanging the NS gene had minimal effect on the virus pathogenicity, and it suggests that other viral genes, or combination of genes, are most likely contributing to the increased virulence of H5N1 HPAI viruses in ducks.

Highly pathogenic (HP) H5N1 avian influenza (AI) is enzootic in several countries of Asia and Africa and constitutes a major threat, at the world level, for both animal and public health. Ducks play an important role in the epidemiology of AI, including HP H5N1 AI. Although vaccination can be a useful tool to control AI, duck vaccination has not proved very efficient in the field, indicating a need to develop new vaccines and a challenge model to evaluate the protection for duck species. Although Muscovy duck is the duck species most often reared in France, the primary duck-producing country in Europe, and is also produced in Asia, it is rarely studied. Our team recently demonstrated a good cross-reactivity with hemagglutinin from clade 2.2 and inferred that this could be a good vaccine candidate for ducks. Two challenges using two French H5N1 HP strains, 1) A/mute swan/France/06299/06 (Swan/06299), clade 2.2.1, and 2) A/mute swan/France/070203/07 (Swan/070203), clade 2.2 (but different from subclade 2.2.1), were performed (each) on 20 Muscovy ducks (including five contacts) inoculated by oculo-nasal route (6 log₁₀ median egg infectious doses per duck). Clinical signs were recorded daily, and cloacal and oropharyngeal swabs were collected throughout the assay. Autopsies were done on all dead ducks, and organs were taken for analyses. Virus was measured by quantitative reverse transcriptase–PCR based on the M gene AI virus. Ducks presented severe nervous signs in both challenges. Swan/070203 strain led to 80% morbidity (12/15 sick ducks) and 73% mortality (11/15 ducks) at 13.5 days postinfection (dpi), whereas Swan/06299 strain produced 100% mortality at 6.5 dpi. Viral RNA load was significantly lower via the cloacal route than via the oropharyngeal route in both trials, presenting a peak in the first challenge at 3.5 dpi and being more stable in the second challenge. The brain was the organ containing the highest viral RNA load in both challenges. Viral RNA load in a given organ was similar or statistically significantly higher in ducks challenged with Swan/06299 strain. Thus, the Swan/06299 strain was more virulent and could be used as a putative challenge model. Moreover, challenged ducks and contacts contained the same amounts of viral RNA load, demonstrating the rapid and efficient transmission of H5N1 HP in Muscovy ducks in our experimental conditions.

Highly pathogenic avian influenza (HPAI) H5N1 virus infections have caused unprecedented morbidity and mortality in different species of domestic and wild birds in Asia, Europe, and Africa. In our previous study, we demonstrated the susceptibility and potential epidemiologic importance of H5N1 HPAI virus infections in Canada geese. In this study, we investigated the potential of preexposure with North American lineage H3N8, H4N6, and H5N2 low pathogenicity avian influenza (LPAI) viruses to cross-protect Canada geese against a lethal H5N1 HPAI virus challenge. Based on our results, birds that were primed and boosted with an H5N2 LPAI virus survived a lethal H5N1 challenge. In contrast, only two of five birds from the H3N8 group and none of the birds preexposed to H4N6 survived a lethal H5N1 challenge. In vitro cell proliferation assays demonstrated that peripheral blood mononuclear cells collected from each group were no better stimulated by homologous vs. heterologous antigens.

In order to investigate viral adaptation mechanisms to poultry, we performed serial in vivo passages of a wild bird low pathogenicity avian influenza isolate of the H7N3 subtype (A/mallard/Italy/33/01) in three different domestic species (chicken, turkey, and Japanese quail). The virus under study was administered via natural routes at the dose of 10⁶ egg infective dose₅₀/0.1 ml to chickens, turkeys, and quails in order to investigate the clinical susceptibility and the shedding levels after infection. Multiple in vivo passages of the virus were performed by serially infecting groups of five naïve birds of each species, with samples collected from a previously infected group. Quails and turkeys were susceptible to infection for 10 serial passages, whereas chickens were susceptible to two cycles of infection only. Infection of chicken with the quail- and turkey-adapted viruses showed an increased pathogenicity and/or shedding, causing more severe clinical signs and/or higher levels of viral excretion compared to the original strain. The data obtained herein suggest that infection of selected avian species may facilitate the adaptation of avian influenza viruses originating from the wild bird reservoir to chicken. This is the first time turkey has been shown to act as a species in which a virus from the wild reservoir can increase its replication activity in other domestic species.

Influenza viruses have a rapid replication cycle, using enzymes without proofreading capacity and generating multiple virus quasispecies during replication. The identification and quantification of these quasispecies populations require time-consuming and expensive cloning and sequencing approaches. In the present study, we developed mutation-specific real-time PCR (RT-PCR) tests for the fine quantification of mutations in a background of wild-type sequences. As a proof-of-concept model, we developed mutation-specific RT-PCR tests to quantify antibody escape mutations during passage under monoclonal antibody (mAb) selection pressure in quasispecies populations of HPAI A/crested eagle/Belgium/01/2004 (H5N1). Mutation-specific RT-PCRs were developed for two mutations (one in HA1 and one in HA2) and validated using plasmids representing either the wild-type sequence or the mutation. The approach achieves a precise and accurate estimation of mutation frequencies on mixed populations in the range of 1% to 99% and does not require standard curves or calibrators. For the HA1 mutations, a directional increase of %G over the passages towards fixation of the G mutation could be observed. On the contrary, as expected from the inaccessibility of the HA2 region to antibodies, the HA2 mutation increased in frequency by factors unrelated to mAb-driven selection. This approach allows in-depth analysis of quasispecies dynamics using large sample sizes. It may also be applied to the dynamics of hot spots of mutations in several genes, such as HA or PB2, and to the early detection of critical changes in the field situation.

H5N1 avian influenza virus has caused widespread infection in poultry and wild birds, and has the potential to emerge as a pandemic threat to humans. The hemagglutinin (HA) is a glycoprotein on the surface of the virus envelope. Understanding its antigenic structure is essential for designing novel vaccines that can inhibit virus infection. The aim of this study was to map the amino acid substitutions that resulted in resistance to neutralization by monoclonal antibodies (MAbs) of the highly pathogenic A/crested eagle/Belgium/01/2004 (H5N1), a clade 1 virus. Two hybridomas specific to H5N1 clade 1 viruses were selected by enzyme-linked immunosorbent assay, virus neutralization test, and immunofluorescence assay. Escape mutant populations resisting neutralization by those MAbs (8C5 and 5A1) were then selected, and sequencing of these mutants allowed the prediction of the HA protein structure by molecular homology. We could detect an amino acid change in our escape mutants at position K189E corresponding to antigenic site 2 of H5 HA1 and site B of H3 HA1. Interestingly, 336 out of 350 available HA sequences from H5N1 clade 1 and clade 2.3 viruses had Lys (K) at position 189 in the HA1, whereas HA sequences analyzed from clade 2.1 and 2.2 viruses had Arg (R). This residue also interacts with the receptor-binding site, and it is thus important for the evolution of H5N1 viruses. An additional substitution K29E in HA2 subunit was also observed and identified with the use of NetChop software as a loss of a proteasomal cleavage site, which seems to be an advantage for H5N1 viruses.

The chicken's major histocompatibility complex (MHC) haplotype has profound influence on the resistance or susceptibility to certain pathogens. For example, the B21 MHC haplotype confers resistance to Marek's disease (MD). However, non-MHC genes are also important in disease resistance. For example, lines 6 and 7 both express the B2 MHC haplotype, but differ in non-MHC genes. Line 6, but not line 7, is highly resistant to tumors induced by the Marek's disease herpesviruses and avian leukosis retroviruses. Recently, survival in the field by Thai indigenous chickens to H5N1 high-pathogenicity avian influenza (HPAI) outbreaks was attributed to the B21 MHC haplotype, whereas the B13 MHC haplotype was associated with high mortality in the field. To determine the influence of the MHC haplotype on HPAI resistance, a series of MHC congenic white leghorn chicken lines (B2, B12, B13, B19, and B21) and lines with different background genes but with the same B2 MHC haplotype (Line 6₃ and 7₁) were intranasally challenged with low dose (10 mean chicken lethal doses) of reverse-genetics–derived rg-A/chicken/Indonesia/7/2003 (H5N1) HPAI virus. None of the lines were completely resistant to lethal effects of the challenge, as evidenced by mortality rates ranging from 40% to 100%. The B21 line had mortality of 40% and 70%, and the B13 line had mortality of 60% and 100% in two separate trials. In addition, the mean death times varied greatly between groups, ranging from 3.7 to 6.9 days, suggesting differences in pathogenesis. The data show that the MHC has some influence on resistance to AI, but less than previously proposed, and non-MHC background genes may have a bigger influence on resistance than the MHC.

Twelve-week-old Vanaraja (an Indian native dual purpose breed) chickens were inoculated intranasally with different doses (100, 1000, and 10,000 mean embryo infective dose [EID50]) of H5N1 virus, and the clinical disease and pathologic changes were compared. Although the overall severity of clinical signs was more severe in the 100 EID50 group, the progression of the clinical disease was slower with delayed onset of mortality when compared with the other two groups. The mean death time of the 100 EID50 group (4.57 days) differed significantly from that of the 10,000 EID50 group (3.60 days) and from that of the 1000 EID50 group (3.33 days). Similarly, overall severity of gross lesions was expressed more in the 100 EID50 group. The histopathologic lesions were of a more hemorrhagic and necrotic nature in the 100 EID50 group, histopathologic lesions were of an inflammatory/proliferative nature in the 1000 EID50 group, and a tendency for intravascular coagulopathy was observed in the 10,000 EID50 group. These differences may be assigned to the influence of dose in the outcome of disease.

Avian influenza virus (AIV) prevalence in wild aquatic bird populations varies with season, geographic location, host species, and age. It is not clear how age at infection affects the extent of viral shedding. To better understand the influence of age at infection on viral shedding of wild bird–origin low pathogenicity avian influenza (LPAI) viruses, mallards (Anas platyrhynchos) of increasing age (2 wk, 1 mo, 2 mo, 3 mo, and 4 mo) were experimentally inoculated via choanal cleft with a 10⁶ median embryo infectious dose (EID₅₀) of either A/Mallard/MN/355779/00 (H5N2) or A/Mallard/MN/199106/99 (H3N8). Exposed birds in all five age groups were infected by both AIV isolates and excreted virus via the oropharynx and cloaca. The 1-month and older groups consistently shed virus from 1 to 4 d post inoculation (dpi), whereas, viral shedding was delayed by 1 d in the 2-wk-old group. Past 4 dpi, viral shedding in all groups varied between individual birds, but virus was isolated from some birds in each group up to 21 dpi when the trial was terminated. The 1-mo-old group had the most productive shedding with a higher number of cloacal swabs that tested positive for virus over the study period and lower cycle threshold values on real-time reverse-transcription PCR. The viral shedding pattern observed in this study suggests that, although mallards from different age groups can become infected and shed LPAI viruses, age at time of infection might have an effect on the extent of viral shedding and thereby impact transmission of LPAI viruses within the wild bird reservoir system. This information may help us better understand the natural history of these viruses, interpret field and experimental data, and plan future experimental trials.

Mutations in a wild duck isolate of avian influenza virus were detected in isolates shed by chickens within 1 day after inoculation. The newly adapted virus was transmitted to naïve chickens in direct contact and sharing food and water. Two consistent amino acid substitutions in the hemagglutinin have been identified, A198V and S274F, and may be important in transmissibility. Mutants with a 30–amino acid deletion in the neuraminidase stalk region 43–72 (N9 numbering) were recovered from inoculated chickens, but not from naïve chickens in contact. The NA stalk mutant virus did not replicate well in Pekin ducks. In vivo viral replication was at low titers and a change in tropism from the respiratory to the digestive tract was observed. Our results indicated that there is a rapid genetic adaptation of wild bird isolates in poultry species, but that resultant viruses may have phenotypes that are intermediate and not fully adapted to the new host.

Diagnosis and management of avian influenza outbreaks now include the use of validated real-time reverse transcription PCR (RRT-PCR) methods in many countries, including all member states of the European Union. Two outbreaks in poultry of notifiable avian influenza (H5 and H7 subtypes) that occurred in Great Britain during 2007 will serve as examples in which RRT-PCR demonstrated its value in 1) rapid diagnosis and confirmation of disease by sensitive and specific laboratory testing of samples derived from the index cases and 2) high-volume, rapid testing of surveillance samples. The two poultry outbreaks followed the incursion of a H7N2 low-pathogenicity notifiable avian influenza (LPNAI) virus (May–June 2007) and a Eurasian lineage H5N1 highly pathogenic notifiable avian influenza (HPNAI) virus (November 2007). Coupled with the use of high-throughput, robotic RNA extraction methods, a total of approximately 9300 and 20,300 field samples were tested by appropriate, validated RRT-PCR assays during the 4- and 5-wk duration of the H7N2 LPNAI and H5N1 HPNAI outbreaks, respectively. Fundamental features of the validated RRT-PCR assays used included their high degree of sensitivity, specificity, and rapidity, attributes that were invaluable in providing timely and accurate information for notifiable AI outbreak management.

This study was aimed at redesigning the Belgian active surveillance program for domestic birds in professional poultry holdings based on a risk analysis approach. A stochastic quantitative analysis, combining all data sources, was run to obtain sensitivity estimates for the detection of an infected bird in the different risk groups identified. An optimal number of holdings for each risk group was then estimated on the basis of the different sensitivities obtained. This study proved to be a useful tool for decision makers, providing insight on how to reallocate the total amount of samples to be taken in the coming year(s) in Belgium, thus optimizing the field resources and improving efficiency of disease surveillance such as required by the international standards.

Effective laboratory methods for identifying avian influenza virus (AIV) in wild bird populations are crucial to understanding the ecology of this pathogen. The standard method has been AIV isolation in chorioallantoic sac (CAS) of specific-pathogen-free embryonating chicken eggs (ECE), but in one study, combined use of yolk-sac (YS) and chorioallantoic membrane inoculation routes increased the number of virus isolations. In addition, cell culture for AIV isolation has been used. Most recently, real-time reverse transcriptase (RRT)-PCR has been used to detect AIV genome in surveillance samples. The purpose of this study was to develop a diagnostic decision tree that would increase AIV isolations from wild bird surveillance samples when using combinations of detection and isolation methods under our laboratory conditions. Attempts to identify AIV for 50 wild bird surveillance samples were accomplished via isolation in ECE using CAS and YS routes of inoculation, and in Madin-Darby canine kidney (MDCK) cells, and by AIV matrix gene detection using RRT-PCR. AIV was isolated from 36% of samples by CAS inoculation and 46% samples by YS inoculation using ECE, isolated from 20% of samples in MDCK cells, and detected in 54% of the samples by RRT-PCR. The AIV was isolated in ECE in 13 samples by both inoculation routes, five additional samples by allantoic, and 10 additional samples by yolk-sac inoculation, increasing the positive isolation of AIV in ECE to 56%. Allantoic inoculation and RRT-PCR detected AIV in 14 samples, with four additional samples by allantoic route alone and 13 additional samples by RRT-PCR. Our data indicate that addition of YS inoculation of ECE will increase isolation of AIV from wild bird surveillance samples. If we exclude the confirmation RT-PCR test, cost analysis for our laboratory indicates that RRT-PCR is an economical choice for screening samples before doing virus isolation in ECE if the AIV frequency is low in the samples. In contrast, isolation in ECE via CAS and YS inoculation routes without prescreening by RRT-PCR was most efficient and cost-effective if the samples had an expected high frequency of AIV.

An indirect enzyme-linked immunosorbent assay (ELISA) was developed using baculovirus, purified, recombinant N1 protein from A/chicken/Indonesia/PA7/2003 (H5N1) virus. The N1-ELISA showed high selectivity for detection of N1 antibodies, with no cross-reactivity with other neuraminidase subtypes, and broad reactivity with sera to N1 subtype isolates from North American and Eurasian lineages. Sensitivity of the N1-ELISA to detect N1 antibodies in turkey sera, collected 3 wk after H1N1 vaccination, was comparable to detection of avian influenza antibodies by the commercial, indirect ELISAs ProFLOK® AIV Plus ELISA Kit (Synbiotics, Kansas City, MO) and Avian Influenza Virus Antibody Test Kit (IDEXX, Westbrook, ME). However, 6 wk after vaccination, the Synbiotics ELISA kit performed better than the N1-ELISA and the IDEXX ELISA kit. An evaluation was made of the ability of the N1-ELISA to discriminate vaccinated chickens from subsequently challenged chickens. Two experiments were conducted, chickens were vaccinated with inactivated H5N2 and H5N9 viruses and challenged with highly pathogenic H5N1 virus, and chickens were vaccinated with recombinant poxvirus vaccine encoding H7 and challenged with highly pathogenic H7N1 virus. Serum samples were collected at 14 days postchallenge and tested by hemagglutination inhibition (HI), quantitative neuraminidase inhibition (NI), and N1-ELISA. At 2 days postchallenge, oropharyngeal swabs were collected for virus isolation (VI) to confirm infection. The N1-ELISA was in fair agreement with VI and HI results. Although the N1-ELISA showed a lower sensitivity than the NI assay, it was demonstrated that detection of N1 antibodies by ELISA was an effective and rapid assay to identify exposure to the challenge virus in vaccinated chickens. Therefore, N1-ELISA can facilitate a vaccination strategy with differentiation of infected from vaccinated animals using a neuraminidase heterologous approach.

Surveillance of wild bird populations for avian influenza viruses (AIV) contributes to our understanding of AIV evolution and ecology. Both real-time reverse transcriptase–polymerase chain reaction (RRT-PCR) and virus isolation in embryonating chicken eggs (ECE) are standard methods for detecting AIV in swab samples from wild birds, but AIV detection rates are higher with RRT-PCR than isolation in ECE. In this study we tested duck embryos, turkey embryos, and multiple cell lines for AIV growth as compared to ECE for improved isolation and propagation of AIV for isolates representing all 16 hemagglutinin subtypes. There were no differences in low pathogenicity AIV (LPAIV) propagation titers in duck or turkey embryos compared to ECE. The replication efficiency of LPAIV was lower in each of the cell lines tested compared to ECE. LPAIV titers were 1–3 log mean tissue-culture infective doses (TCID₅₀) lower in Madin-Darby canine kidney (MDCK), primary chicken embryo kidney (CEK), and primary chicken embryo fibroblast (CEF) cell cultures. and 3–5 log TCID₅₀ lower in chicken bone marrow macrophage (HD11), chicken fibroblast (DF-1), and mink lung epithelial (Mv1Lu) cells than the corresponding mean embryo infective doses (EID₅₀) in ECE. The quail fibroblast (QT-35) and baby hamster kidney (BHK-21) cell lines produced titers 5–7 log TCID₅₀ less than EID₅₀ in ECE. Overall, ECEs were the most efficient system for growth of LPAIV. However, the savings in time and resources incurred with the use of the MDCK, CEK, and CEF cultures would allow a higher volume of samples to be processed with the same fiscal and financial resources, thus being potentially advantageous despite the lower replication efficiency and lower isolation rates.

Since the emergence of the highly pathogenic avian influenza H5N1, avian influenza surveillance has been expanded in Europe. The serologic monitoring of domestic poultry is usually accomplished using the reference hemagglutination inhibition (HI) test for the detection of H5 and H7 subtypes. However, as the number of tested sera has been increasing, there is a need for another serologic method that could be used as a preliminary screening test. A comparison of four enzyme-linked immunosorbent assay (ELISA) tests (two indirect and two competitive) was conducted, and they showed good specificity and higher sensitivity than the HI test. The selected ELISA tests were then tested using approximately 800 field sera representative of different poultry species, and a simulation was done to determine the best strategy for screening. The first strategy was testing both gallinaceous and nongallinaceous sera with a competitive ELISA and using the HI test for H5 and H7 as a confirmatory test. The second strategy was testing only gallinaceous bird sera with the indirect ELISA with confirmatory H5 and H7 HI and all nongallinaceous sera by the H5 and H7 HI test. In the Belgian poultry context, the best strategy seems to be the use of a blocking ELISA as the primary screening tool to test all the poultry sera, followed by confirmation by H5 and H7 HI test subtyping.

A real-time reverse transcription PCR (RRT-PCR) targeting a highly conserved HA2 H7 region was developed for the detection of all H7 subtype avian influenza viruses (PanH7). The wide phylogenetic scope and analytical sensitivity and specificity were validated with the use of a panel of 56 diverse influenza A viruses. The detection limit was determined with the use of serial dilutions of Eurasian isolates A/Ck/BE/06775/2003 and A/Ck/It/1067/v99 and North American isolates A/CK/PA/143586/2001 and A/Quail/PA/20304/1998, to be 1 log₁₀ higher than the detection limit of the generic influenza A matrix RRT-PCR (about 2.5 EID₅₀/reaction compared to 0.25 EID₅₀/reaction for matrix). Diagnostic test properties of PanH7 were determined with the use of 102 swabs from A/Ck/It/1067/v99 experimentally infected chickens, and were not affected by the increased detection limit of PanH7. In comparison to matrix RRT-PCR and virus isolation in embryonated chicken eggs (VI), the PanH7 detected more weakly positive oropharyngeal swabs at the onset of the infection. PanH7 diagnostic sensitivity compared to virus isolation (VI) was 83.3% (compared to 72.2% for matrix RRT-PCR); and diagnostic specificity was 88.1% (94.0% for matrix). The PanH7 test can also be tailored to detect only American (AmH7) or only Eurasian (EurH7) strains by changing the mix of forward and reverse primers used in combination with the unique probe. Overall, this new test is a valuable tool for the detection and identification of H7 subtype influenza A.

This report describes the validation of an avian influenza virus (AIV) H7 subtype-specific real-time reverse transcriptase–PCR (rRT-PCR) assay developed at the Southeast Poultry Research Laboratory (SEPRL) for the detection of H7 AI in North and South American wild aquatic birds and poultry. The validation was a collaborative effort by the SEPRL and the National Veterinary Services Laboratories. The 2008 H7 rRT-PCR assay detects 10¹ 50% embryo infectious doses per reaction, or 10³–10⁴ copies of transcribed H7 RNA. Diagnostic sensitivity and specificity were estimated to be 97.5% and 82.4%, respectively; the assay was shown to be specific for H7 AI when tested with >270 wild birds and poultry viruses. Following validation, the 2008 H7 rRT-PCR procedure was adopted as an official U.S. Department of Agriculture procedure for the detection of H7 AIV. The 2008 H7 assay replaced the previously used (2002) assay, which does not detect H7 viruses currently circulating in wild birds in North and South America.

The hemagglutinin gene of an avian influenza virus (AIV) A/duck/NC/674964/07 (H5N2) was cloned and expressed in a baculovirus system (H5-Bac). In parallel, a recombinant hemagglutinin of A/Vietnam/1203/04 (H5N1) was expressed in mammalian cells, purified, and used for generation of H5-specific monoclonal antibodies (MAb). The purified H5-Bac was used to develop a competitive enzyme-linked immunosorbent assay (cELISA) to detect H5 antibodies in a species-independent approach using one of the established H5-specific MAbs as the competitor antibody. The cELISA performed with influenza antibody-free sera or with sera of animals infected with other than H5-encoding AIV showed no significant inhibition of H5-MAb binding, indicating high test specificity. In contrast, sera of poultry (chickens, turkeys, ducks) experimentally infected with H5-encoding AIV were able to significantly inhibit the binding of the MAb in a species-independent approach. Comparison of the results of the cELISA with results obtained by a hemagglutination inhibition assay showed a gradient of the sensitivity (turkeys > ducks > chicken). The described results show that H5-specific antibodies in sera can be detected in a species-independent approach by using a recombinant protein.

Early detection of highly pathogenic (HP) strains of avian influenza, especially the HP H5N1, is important in terms of controlling and minimizing the spread of the virus. Several rapid antigen detection kits that are able to detect influenza A viruses in less than 1 hr are commercially available, but only a few of them have been evaluated. In this study, four commercially available rapid tests for veterinary usage and two tests for human usage were evaluated and compared. The evaluation of the detection limits of the different tests established with serial dilution of HP H5N1 indicated that most of them have a detection limit of about 10⁵ to 10⁶ 50% tissue culture infectious dose/ml. None of the tests was able to detect virus in oral and cloacal swabs 24 hr post–experimental infection of specific-pathogen-free chickens with HP H5N1. However, 48 hr postinfection, almost all of the rapid tests were able to detect infected birds (dead or alive). Moreover, organs were also successful samples for detection of H5N1 with the rapid tests. Unexpectedly, the specificity was not very high for some tests. However, in general in this study, the tests for veterinary usage showed better sensitivity. To conclude, these tests offer good indicative value in the event of an outbreak, but as a result of their low sensitivity and some aspecific reactions, test results always need to be confirmed by other methods.

Historically, virus isolation has been the method of choice for conducting surveillance for avian influenza virus (AIV) in avian species. More recently, the primary screening method has become real-time reverse transcription–polymerase chain reaction (RRT-PCR). We wanted to determine how these two testing methods (virus isolation and RRT-PCR) affected AIV prevalence estimation, particularly in an understudied, low-prevalence region—the waterfowl wintering grounds along the Texas mid–Gulf Coast. Cloacal swabs were collected from hunter-harvested waterfowl and other wetland-associated game birds during four consecutive hunting seasons (2005–2006 through 2008–2009). Overall prevalence by RRT-PCR (5.9%, 6.5%, 11.2%, and 5.5%) was approximately an order of magnitude higher than prevalence by virus isolation (0.5%, 1.3%, 3.9%, and 0.7%) for the four hunting seasons, respectively. Apparent AIV prevalence by virus isolation conducted only on RRT-PCR–positive samples resulted in estimates nearly identical in magnitude to those derived from parallel testing (0.5% vs. 0.6%, 1.3% vs. 1.7%, and 3.9% vs. 4.0% for 2005–2006, 2006–2007, and 2007–2008, respectively). Unlike most reports of seasonal variation in AIV prevalence, we documented differences in prevalence estimates among months by RRT-PCR only during 2008–2009 and by virus isolation only during 2006–2007 and 2007–2008. Our data indicate that screening samples by RRT-PCR followed by virus isolation only on RRT-PCR–positive samples provides a reasonable means to generate prevalence estimates close to the true prevalence as determined by virus isolation. We also confirmed the low prevalence of AIV in waterfowl wintering grounds along the Texas mid–Gulf Coast and demonstrated little variation in prevalence among months during the four hunting seasons sampled.

The role of wild ducks as vectors of avian influenza viruses (AIVs) is well known but the immune response induced by AIV has different patterns according to the species of duck. The local antibody produced on the mucosal surface may play an important role not only in protection but also in limitation of the primary replication at the portal of entry and shedding of the virus. With this aim, specific enzyme-linked immunosorbent assays (ELISAs) for duck IgY, Ig light chain, IgY (heavy chain), IgA, and IgM were developed and used to evaluate the systemic and mucosal response induced in ducks after low pathogenic avian influenza (LPAI) infections. Two different species of ducks (Mule and Pekin), ages (1 wk and 3 wk), and virus strains (H7N1 and H5N1 low pathogenic viruses) were tested in two studies to evaluate the developed tools. In the two studies, systemic and mucosal AIV-specific duck IgY, Ig (light chain), IgY (heavy chain), IgA, and IgM were detected and followed. Therefore, the developed ELISAs proved to be efficient tools allowing the follow-up of the systemic and mucosal responses induced by LPAI infection in ducklings. These tools can be very helpful for the development and evaluation of AI vaccines for ducks.

Avian influenza (AI) surveillance in commercial poultry is accomplished by detecting the presence of antibodies to two group-specific antigens, NP and M1, using the agar gel immunodiffusion test. In order to determine the viral subtype responsible for the infection, positive samples must be further subtyped using the hemagglutination inhibition and neuraminidase inhibition tests. These tests are labor intensive and may take up to 4 days, thus slowing down responses to outbreaks. To expedite the subtyping of chicken sera we have developed a multiplex fluorescence microsphere immunoassay (FMIA), which allows for the simultaneous detection and subtyping of chicken sera to H5 influenza viruses. The FMIA was developed using NP (full length) and H5 (HA1 region) proteins expressed in baby hamster kidney cells using a Venezuela equine encephalitis virus replicon system. Both proteins were tagged with 6xHis at the carboxy-end and purified using cobalt-coated agarose beads. Purified H5 protein showed minimal cross-reactivity with anti-H2 serum, while no cross-reactivity was observed with sera to other AI virus (AIV) subtypes and other important poultry viral pathogens. In addition, and as expected, all the AIV sera tested reacted strongly with purified NP protein. Our results indicate that FMIA can be used for rapid subtyping of chicken sera.

The highly pathogenic avian influenza virus of subtype H5N1 that caused serious outbreaks in Egypt in 2006 was efficiently detected using a commercially available real-time reverse transcriptase–PCR (RRT-PCR) for the type A specific matrix (M) gene in field samples of cloacal and tracheal swabs. RRT-PCR was also used for subtyping and confirmation of H5 subtype. During late 2007 the National Laboratory for Veterinary Quality Control on Poultry Production detected five field cases that were positive for avian influenza virus (AIV) based on the M gene RRT-PCR. Three different commercial H5 RRT-PCRs were used for identification of the H5 subtype, as well as a published World Organization for Animal Health (OIE) H5 RRT-PCR that had been previously carefully validated. The five cases had positive results for the H5 gene using the published OIE H5 RRT-PCR, but the three commercial H5 RRT-PCRs tests only returned two to four positive results out of the five positive cases. The hemagglutinin gene (HA) sequencing analysis of these five isolates showed multiple nucleotide substitution mutations, suggesting genetic variation that could affect the H5 primer and/or probe binding sequences. These data highlight the importance of continued monitoring of RRT-PCR primers and probes to ensure that sensitivity and specificity are maintained. The use of conventional methods in national and reference AIV laboratories, including virus isolation, serologic subtyping, and alternative RRT-PCR primers, is necessary to detect the newly emerging variant H5N1 strains that affect diagnostic performance.

In a previous study, we optimized DNA barcoding techniques for avian influenza virus (AIV) isolation and host identification, using fecal samples from wild birds, for high-throughput surveillance of migratory waterfowls. In the present study, we surveyed AIV in Mongolia during the breeding season and, subsequently, in Korea in winter, to compare prevalent AIV subtypes and hosts using DNA barcoding. In Korea, H4 and H5 subtypes were the most abundantly detected HA subtypes, and most AIVs were isolated from the major population (mallards, Anas platyrhynchos) of wild bird habitats. On the other hand, in Mongolia, H3 and H4 subtypes were the most abundantly detected HA subtypes, and most AIVs were isolated from a small population of wild bird habitats that were not visible at the sampling site. In conclusion, AIV isolation using fecal samples, accompanied with DNA barcoding techniques as a host bird species identification tool, could be useful for monitoring major and minor populations of wild bird habitats. Further, continuous, and large-scale surveillance could be helpful for understanding the AIV epidemiology, evolution, and ecology in wild waterfowl.

A competitive enzyme-linked immunosorbent assay (c-ELISA) was developed as a serologic diagnostic tool to detect antibodies against NA subtype 3 of avian influenza virus (AIV). The NA antigen used in this c-ELISA was obtained by pronase treatment of allantoic fluid of specific-pathogen-free (SPF) eggs infected with AIV. The NA specific monoclonal antibodies were produced from purified NA. The N3 c-ELISA was carried out on serum samples collected from both SPF chickens and commercial layers to confirm whether the N3 c-ELISA was capable of detecting specific N3 antibodies. The positive cutoff percentage inhibition value was 6.13%. The sensitivity and specificity of the N3 c-ELISA were 83.7% and 95.6%, respectively, which indicated that N3 c-ELISA can detect the antibodies from SPF chickens or commercial chickens vaccinated with H9N3 subtype of AIV.

New lyophilized real-time reverse transcription (RT)-PCR avian influenza detection assays were designed and tested. The M-gene assay detects all avian influenza virus (AIV) subtypes, and the H5 and H7 specific assays can discriminate the AIV subtypes H5 and H7 of Eurasian origin. The assays are formulated in a lyophilized bead format containing an internal positive control to monitor inhibitors in the reaction. Fifty-six AIV cultured isolates covering all 16 hemagglutinin types and 44 positive swabs from an outbreak of AIV in turkeys (H5N1 highly pathogenic avian influenza ) were used to determine analytical performance and diagnostic sensitivity of these veterinary assays. The lyophilized real-time RT-PCR assays were demonstrated to be more sensitive than the wet assays, being able to detect down to 4 to 16 molecules of synthetic target RNA compared to 16 to 80 molecules for the corresponding wet assays. The diagnostic sensitivity of the lyophilized M-gene assay was determined to be 97.7% (43/44), whereas concurrent testing of these samples with the wet assay was only 86.3% sensitive (38/44). Using a panel of 19 noninfluenza respiratory and enteric pathogens, the analytical specificity of the M-gene assay was shown to be 100%. High diagnostic specificity of the assays was also confirmed by testing 496 negative swab samples from a combination of wild bird species and poultry.

Viruses of all subtypes may be introduced into domestic poultry, but only H5 and H7 low pathogenicity avian influenza viruses can mutate during circulation in poultry and emerge as high pathogenicity avian influenza variants. It is therefore essential to monitor the field situation continuously to detect low pathogenicity notifiable avian influenza (LPNAI) as soon as possible. With the emergence of the highly pathogenic H5N1, markers of infection of avian influenza are becoming more and more significant. They are important as early warning systems, and they can also be valuable tools as companion tests of vaccines (differentiating infected from vaccinated animals), but they might also be informative about the evaluation of the degree of adaptation and the timing of infection. Therefore, several experimental infections of specific-pathogen-free chickens were conducted to follow the kinetics of antibody responses against the hemagglutinin, the neuraminidase, the nucleoprotein, and the M2e after infections with LPNAI viruses isolated from waterfowl or already adapted to chicken. Overall, the immune responses against the different antigens showed similar kinetics in the different infected animals, but they were lower when the animals were infected with AI viruses originating from waterfowl, and the kinetics of the M2e antibodies was quite different. Indeed, it was rather shorter and disappeared more rapidly (approximately 35 days postinfection) compared to the kinetics of the other antibodies. Therefore, the detection of the antibodies against M2e peptide could be an interesting tool to detect recent infection, and these preliminary results indicated that the production of M2e antibodies might be correlated with the degree of adaptation of LPNAIs.

Highly pathogenic avian influenza (HPAI) strain H5N1 has received great attention with regard to its potential spread to North America. This quantitative risk assessment, which is primarily based on wild bird carriage of HPAI from East Asia to Alaska, was conducted to assess the likelihood of a hunter retriever dog becoming infected after harvesting an infected waterfowl during the Alaskan hunting season. Using Monte Carlo Simulation with @Risk software, the expected probability of a hunter retriever dog becoming infected is 2.3 × 10⁻⁸. This model can serve as a tool for decision makers in assessing the risk of HPAI strain H5N1 introduction into Alaska's hunter retriever dogs.

In the past decades, mathematical models have become more and more accepted as a tool to develop surveillance programs and to evaluate the efficacy of intervention measures for the control of infectious diseases such as highly pathogenic avian influenza. Predictive models are used to simulate the effect of various control measures on the course of an epidemic; analytical models are used to analyze data from outbreaks or from experiments. A key parameter in both types of models is the reproductive ratio, which indicates whether virus can be transmitted in the population, resulting in an epidemic, or not. Parameters obtained from real data using the analytical models can subsequently be used in predictive models to evaluate control strategies or surveillance programs. Examples of the use of these models are described here.

The potential spread of highly pathogenic avian influenza among commercial broiler farms in Georgia, U. S. A., was mathematically modeled. The dynamics of the spread within the first infected flock were estimated using an SEIR (susceptible-exposed-infectious-recovered) deterministic model, and predicted that grower detection of flock infection is most likely 5 days after virus introduction. Off-farm spread of virus was estimated stochastically for this period, predicting a mean range of exposed farms from 0–5, depending on the density of farms in the area. Modeled off-farm spread was most frequently associated with feed trucks (highest daily probability and number of farm visits) and with company personnel or hired help (highest level of bird contact).

The tenacity of three low pathogenicity avian influenza viruses (AIV; subtypes H4N6, H5N1, and H6N8) was tested at five different temperatures (−10, 0, 10, 20, and 30 C) in distilled water, normal saline, and surface water obtained from Lake Constance. Infectivity of AIV in the samples was quantified at regular intervals by end point titration on Madin-Darby canine kidney cells for a maximum period of 36 wk, and duplicate samples were tested each time. The results showed that the survival time of AIV in all of the water types was inversely proportional to storage temperature. All three viruses showed varying sensitivity to inactivation under each of the experimental conditions. Persistence of the viruses was the longest in distilled water, second longest in normal saline, and shortest in surface water. The virus-inoculated surface water remained infective for a few days at 30 and 20 C, a few weeks at 10 C, and for months at 0 and −10 C.

The H9N2 avian influenza virus (AIV) subtype has become endemic in Israel since its introduction in 2000. The disease has been economically damaging to the commercial poultry industry, in part because of the synergistic pathology of coinfection with other viral and/or bacterial pathogens. Avian influenza virus viability in the environment depends on the cumulative effects of chemical and physical factors, such as humidity, temperature, pH, salinity, and organic compounds, as well as differences in the virus itself. We sought to analyze the viability of AIV H9N2 strains at three temperatures (37, 20, and 4 C) and at 2 pHs (5.0 and 7.0). Our findings indicated that at 37 C AIV H9N2 isolate 1525 (subgroup IV) survived for a period of time 18 times shorter at 20 C, and 70 times shorter period at 4 C, as measured by a decrease in titer. In addition, the virus was sensitive to a lower pH (pH 5.0) with no detectable virus after 1 wk incubation at 20 C as compared to virus at pH 7.0, which was viable for at least 3 wk at that temperature. The temperature sensitivity of the virus corresponds to the occurrence of H9N2 outbreaks during the winter, and lower pH can greatly affect the viability of the virus.

Serologic testing of wild birds for avian influenza virus (AIV) surveillance poses problems due to species differences and nonspecific inhibitors that may be present in sera of wild birds. Recently available competitive enzyme-linked immunosorbent assay (cELISA) kits offer a new species-independent approach. In this study we compare two commercial competitive cELISAs, using a total of 184 serum and plasma samples from 23 species of wild birds belonging to 10 orders. Thirteen samples were from experimentally high pathogenicity AI and low pathogenicity AI infected red-legged partridges (Alectoris rufa), 77 samples were from a flock of sentinel hybrid ducks confirmed infected by AI by real-time PCR, and 94 samples were from wild birds admitted to a rehabilitation center. Both ELISAs detected AI antibodies in the experimentally infected partridges, whereas hemagglutination inhibition (HI) was negative. Concordance in results between the two ELISAs was 51.5%. When specific subtype-H5/H7 HI-positive samples were considered for comparison, ELISA 1 appeared to perform better on ducks, whereas ELISA 2 appeared to perform better in other wild bird species. Overall, 68.2% of H5/H7 positive samples tested positive by ELISA 1 and 36% by ELISA 2. Both ELISAs detected AIV-antibody–positive samples negative by specific HI against 9 of the 16 existing hemagglutinin (HA) subtypes. Presumably this reflects either higher sensitivity of cELISA when compared to HI, presence of antibodies against HA subtypes not tested, or unspecific reactions. Performance of ELISA 1 on ducks appears to be comparable to in-house cELISA previously used by other authors in wild birds, but requires a relatively large sample volume. Alternatively, although ELISA 2 required a smaller sample volume, it was less effective at identifying HI-positive samples. The results reflect the necessity of validation of cELISA tests for individual species or at least families, as required by the OIE.

Persistence of H5N1 high pathogenicity avian influenza virus (HPAIV), isolated during the epidemic in wild birds in Poland in 2006, was evaluated in three water samples derived from the sources known to host wild water birds (city pond, Vistula river mouth, and Baltic Sea). The virus was tested at two concentrations (10⁴ and 10⁶ median tissue culture infective dose per milliliter) and at three temperatures (4 C, 10 C, and 20 C), representing average seasonal temperatures in Poland. All tested water samples were filtered before virus inoculation, and one unfiltered sample (Baltic seawater) was also tested. Infectivity was determined twice a week over a 60-day trial period by microtiter endpoint titration. The persistence of the virus varied considerably depending on its concentration and also on physico-chemical parameters of the water, such as temperature and salinity. Avian influenza virus survival was the highest at 4 C and the lowest at 20 C. Prolonged infectivity of the virus in Baltic seawater (brackish, 7.8 ppt) was also seen. In distilled water, the virus retained its infectivity beyond the 60-day study period. Interestingly, a devastating effect of the unfiltered fraction of seawater was seen as the virus disappeared in this fraction the quickest in all studied combinations; thus, biologic factors may also affect infectivity of HPAIV.

Live bird markets (LBMs) provide an ideal environment for the evolution and interspecies transfer of avian influenza viruses (AIVs). In this study, we analyzed AIVs present in LBMs in Korea during the winter seasons of 2006–08. Sixty-five AIVs that belong to four hemagglutination (HA) subtypes of AIV (H3, H4, H6, and H9) were isolated from 644 pooled tissue or swab samples collected in LBMs. Most H9 subtypes of AIVs were isolated from Galliformes (chickens, silky fowls, pheasants, and guinea fowls), and other subtypes were isolated from Anseriformes (Pekin ducks and mallards). In addition, we obtained a single H3N2 virus from nasal swabs of dogs sold in LBMs, and the virus was genetically identical to the canine influenza virus (CIV) isolated from pet dogs in Korea. Phylogenetic analysis suggests that the Korean H9N2 viruses prevalent in chickens have provided their gene segments to AIVs circulating in ducks. These gene transfers facilitated reassortment events among AIVs and likely generated the ancestors of CIV in Korea. An animal challenge study using chickens, quail, mice, and dogs had shown that the H4 and H6 subtypes could replicate in mice and that some H4 and H6 viruses could replicate in chickens without preadaptation. In addition, two H3 subtype viruses (H3N2 and H3N8) induced interstitial pneumonia that accompanied clinical signs and seroconversion in dogs. Our findings indicate that the newly evolved AIVs have been continuously generated by reassortment in ducks, and these reassortments could result in expanding the host range of AIVs.

The participatory disease surveillance and response (PDSR) approach to highly pathogenic avian influenza (HPAI) in Indonesia has evolved significantly from the participatory disease surveillance (PDS) system developed for rinderpest eradication in Africa and Pakistan. The first phase of the PDSR project emphasized the detection and control of HPAI by separate PDS and participatory disease response teams primarily in sector 4 poultry at the household level. Lessons learned during the first phase were taken into account in the design of the second phase of the project, which has sought to further strengthen management of disease prevention and control activities by improving technical approaches, increasing active participation of key stakeholders, including local and central governments, and focusing on the village level. The ongoing evolution of the PDSR program aims to establish a sustainable community-based program within provincial and district livestock services that enhances the prevention and control of not only HPAI, but also other zoonotic and priority animal diseases.

The implementation of strategies to detect, prevent, and control highly pathogenic avian influenza (HPAI) in developing countries presents several challenges, one of which is the presence of other diseases in poultry populations. Training workshops in developing countries using the Avian Flu School have revealed that in areas with heavy Newcastle disease burdens, smallholder poultry keepers do not recognize HPAI as an immediate threat. We have developed a strategy to address the more proximal needs and priorities of communities with free-ranging poultry flocks as a means to create value in poultry, and thus to improve disease detection and prevention overall. To this end, we have created the Poultry Health and Well-Being for Development project, which trains graduate veterinarians and paraprofessionals in poultry disease diagnosis, control, and treatment. These trainees then serve their local communities to improve poultry health and to implement disease detection and management programs.

Current control strategies for avian influenza (AI) and other highly contagious poultry diseases include surveillance, quarantine, depopulation, disposal, and decontamination. Selection of the best method of emergency mass depopulation involves maximizing human health and safety while minimizing disease spread and animal welfare concerns. Proper selection must ensure that the method is compatible with the species, age, housing type, and disposal options. No one single method is appropriate for all situations. Gassing is one of the accepted methods for euthanatizing poultry. Whole-house, partial-house, or containerized gassing procedures are currently used. The use of water-based foam was developed for emergency mass depopulation and was conditionally approved by the United States Department of Agriculture in 2006. Research has been done comparing these different methods; parameters such as time to brain death, consistency of time to brain death, and pretreatment and posttreatment corticosterone stress levels were considered. In Europe, the use of foam with carbon dioxide is preferred over conventional water-based foam. A recent experiment comparing CO₂ gas, foam with CO₂ gas, and foam without CO₂ gas depopulation methods was conducted with the use of electroencephalometry results. Foam was as consistent as CO₂ gassing and more consistent than argon-CO₂ gassing. There were no statistically significant differences between foam methods.

The chemical compound metam-sodium was tested at three concentrations for the ability to inactivate the infectivity of low-pathogenic avian influenza virus (LPAIV) and infectious bursal disease virus (IBDV), with virus-contaminated chicken litter used as the substrate. LPAIV was inactivated within 1 hr after the addition of metam-sodium independent of the concentration used. IBDV was not inactivated with the lowest amount of metam-sodium, but at higher concentrations the virus was inactivated within 1 hr after application. The results show that metam-sodium is able to penetrate chicken litter and inactivate enveloped as well as nonenveloped viruses because of its ability to form the active compound methyl isothiocyanate, which acts as a fumigant.

Free-ranging local chicken flocks are important for the livelihood of resource-poor rural farmers in Tanzania, as they provide a critical source of animal protein and a ready source of income through the sale of chickens and eggs. An occurrence of highly pathogenic avian influenza (HPAI) in the village setting of Tanzania would result in a disastrous loss of livelihood. This paper attempts to offer an alternative method for preventing and controlling HPAI in village settings of Tanzania through community-based approaches.

Current control strategies for avian influenza virus, exotic Newcastle disease, and other highly contagious poultry diseases include surveillance, quarantine, depopulation, disposal, and decontamination. Skid steer loaders and other mobile equipment are extensively used during depopulation and disposal. Movement of contaminated equipment has been implicated in the spread of disease in previous outbreaks. One approach to equipment decontamination is to power wash the equipment, treat with a liquid disinfectant, change any removable filters, and let it sit idle for several days. In this project, multiple disinfectant strategies were individually evaluated for their effectiveness at inactivating Newcastle disease virus (NDV) on mechanical equipment seeded with the virus. A small gasoline engine was used to simulate typical mechanical equipment. A high titer of LaSota strain, NDV was applied and dried onto a series of metal coupons. The coupons were then placed on both interior and exterior surfaces of the engine. Liquid disinfectants that had been effective in the laboratory were not as effective at disinfecting the engine under field conditions. Indirect thermal fog showed a decrease in overall virus titer or strength. Direct thermal fog was more effective than liquid spray application or indirect thermal fog application.

The response to an avian influenza outbreak, especially highly pathogenic avian influenza (HPAI), should focus on four basic principles: 1) protect humans first, 2) protect animals 3) contain the virus, and 4) ensure that the outbreak remains a single event, by preparing response teams to work together effectively through advance training. The Stamping Out Pandemic and Avian Influenza (STOP AI) project is a U.S. Agency for International Development–funded global activity. STOP AI has designed and conducted practical, experiential training exercises that engage participants in simulated experiences that enhance their confidence and ability to apply these principles during a real HPAI outbreak. This article describes three specific exercises: 1) wearing and removing personal protective equipment (PPE) in a controlled environment, 2) site zoning, and 3) a planning and resource mapping exercise staged in Poultopia—a fictional region in a developing country. The PPE activity emphasizes the physical challenges of working in full PPE and the importance of proper equipment removal. In the zoning exercise, response teams focus on the areas and the tasks required by setting up a clean area, transit corridor, infected/culling area, and nontransit areas at a farm, village, or other location. In Poultopia, participants must determine where surveillance should occur, decide where roadblocks should be placed during an outbreak, choose which birds to cull first and determine how to dispose of them safely, ascertain the types of personnel and equipment needed, and assess timing issues. The Poultopia scenarios are adapted to the conditions of the region where the training takes place, thus adding to their realism and utility. The practical techniques described here have been taught successfully through STOP AI in more than 30 countries in Europe, Asia, and Africa.

The success of emergency intervention to control contagious animal diseases is dependent on the preparedness of veterinary services. In the framework of avian influenza preparedness, the Italian Ministry of Health, in cooperation with the National Reference Centers for Epidemiology and Avian Influenza, implemented an electronic learning course using new web-based information and communication technologies. The course was designed to train veterinary officers involved in disease outbreak management, laboratory diagnosis, and policy making. The “blended learning model” was applied, involving participants in tutor-supported self-learning, collaborative learning activities, and virtual classes. The course duration was 16 hr spread over a 4-wk period. Six editions were implemented for 705 participants. All participants completed the evaluation assignments, and the drop out rate was very low (only 4%). This project increased the number of professionals receiving high-quality training on AI in Italy, while reducing expenditure and maximizing return on effort.