Abstract

Avian influenza (AI) viruses have evolved little in 60 yr in their natural hosts (wild waterfowl, shorebirds and gulls) that nearly always remain asymptomatic. They evolve much faster in aberrant hosts (pigs, chickens) in which they can become lethal. Although often blamed, wild birds are not significant in spreading virulent AI. Responses to AI outbreaks must be based on facts and not media hype. Poultry present the real danger.

Introduction

The two surface glycoproteins of influenza A viruses, haemagglutinin (H) and neuraminidase (N), are the most important for inducing immunity and therefore vary the most.² Few H and N subtypes have been isolated from mammals, but all 15 H and all 9 N subtypes have been isolated from migrating waterfowl and shorebirds worldwide. Natural influenza A infections have been reported in humans, pigs, horses, marine mammals, mustelids and birds.^23,3

Although influenza viruses infect a wide variety of birds and mammals, the natural hosts are wild ducks, geese, swans, gulls and terns that intermittently transmit AI to other avian and even mammalian species (chickens, turkeys, pigs, horses, seals, whales, humans).^18,5 Avian influenza viruses evolve slowly in their natural hosts because of the brief avian lifespan and replication in their intestines and may be in evolutionary stasis.⁸ The evolutionary rate accelerates rapidly in new (aberrant) host species due to selective pressures to adapt.¹⁹

Avian influenza is mostly asymptomatic in aquatic birds.^18,21 Viral replication in aberrant hosts is usually limited and overt disease is rare.⁵ AI viruses bind preferentially to SAa2,3-galactose. Human strains preferentially bind to SAa2,6-galactose.²⁴ Thus, AI viruses do not replicate well in humans, and must reassort or adapt in an intermediate aberrant host before emerging in human populations. Pigs have receptors for both avian and human influenza viruses and are a likely intermediate host. The recent transmissions of avian H5N1 and H9N2 viruses directly to humans showed poultry can also be intermediate hosts.²¹

The first known direct transmission of virulent avian influenza (Hong Kong H5N1) from poultry to people killed 6 of 18 infected people in 1997 (thousands more people were exposed to infected chickens).²⁴ Human-to-human transmission was rare,⁷ and no more human cases occurred after all poultry in Hong Kong were culled. H5N1 was evolving rapidly in the new chicken host and had acquired a number of amino acids that correlate with replication in humans. Eradicating 1.6 million chickens eliminated the immediate opportunity for H5N1 viruses to infect humans.²⁴

Before depopulation, H5N1 virus was isolated from 20% of chickens and 5% of waterfowl in Hong Kong markets, but was not isolated from other birds, including other gallinaceous species, pigeons, and caged passerine and psittacine birds, or from wild birds. Chickens were the only clinically affected species in the live markets.¹² It is likely that H5N1 viruses are now widespread around Hong Kong. The multiplicity of H5N1 genotypes circulating in poultry in the wider region increases the opportunity for the emergence of pandemic strains by developing efficient human-to-human transmission through further reassortment.⁵ Prior to 2003, wild ducks were not found to maintain virulent H5 influenza viruses.¹⁴ H5 viruses can become highly pathogenic in domestic poultry but usually remain non-pathogenic in ducks.¹

Avian influenza viruses are rarely isolated from passerines or psittacines. However, passerines are common near intensive poultry production worldwide and limited evidence supports the potential perpetuation and transmission of AI by passerines near intensive poultry production.¹² Most AI viruses from psittacines have been isolated during quarantine. The H9N2 viruses isolated from two ring-necked parakeets imported from Pakistan into Japan shared high sequence similarities with the 1997 H5N1 and 1999 H9N2 viruses transmitted directly from birds to humans.¹² The two H9N2 isolates identified 1 yr apart were closely related, indicating they belong to the same lineage that must have been established in Pakistan for at least 1 yr. These isolates were non-pathogenic in chickens and mice.¹⁰ Although psittacines are not significant in the epidemiology of influenza A viruses, they can harbour and possibly transmit AI. This risk is greatest in countries with local, regional and international trading of wild birds.¹²

When tested with the 1997 H5N1 virus, seven gallinaceous spp. (chicken, turkey, Japanese quail, bobwhite quail, pearl guinea fowl, ringneck pheasant and chukar partridge) and zebra finches were the most susceptible (high morbidity, mortality >75%) with high viral re-isolation. Geese, emus, house finches and budgerigars were less susceptible and virus re-isolation was low.¹² Ducks, house sparrows and gulls showed mild or no disease, and viral re-isolation was low to moderate. Pigeons, starlings, rats and rabbits resisted infection. This contrasts with previous experiments in which H7N7 virus killed all starlings and spread to contact starlings, but killed only 30% of sparrows and failed to spread to contact sparrows.¹¹ Thus, although the virulence of a single AI virus can vary substantially between avian species, including species within the same order, passerines and psittacines appear to play very minor roles in the natural epidemiology of AI.¹²

The pattern of spread of the virulent 2003/2004 H5N1 outbreak strongly suggests the virus was carried by smuggled poultry, a practice widespread in Southeast Asia. The genetic sequence of the virus isolated from a Vietnamese victim matched most closely one from Chinese poultry. Five of the eight genetic strands were almost identical to an H5N1 from duck meat smuggled from eastern China to Taiwan in 2003.¹³ Some experts blamed migratory birds, but there is no direct evidence of wild birds spreading virulent AI. Wild birds were affected near big poultry outbreaks but regular monitoring of migratory birds in Thailand and elsewhere did not reveal the virus.¹³

The genetic diversity of AI viruses circulating in poultry in southeastern China has increased sharply since 2001. This shows H5 is circulating widely somewhere, under unusual selective pressures.²² Asia’s growing prosperity has caused a boom in intensive poultry production. Since 1997, many Chinese producers have vaccinated with inactivated H5N1. If a vaccine is a poor match, as is the case with the H5N1 strain that swept Asia, AI can still replicate in animals that show no disease. Intensive vaccination in south China (>11x10⁶) may have allowed the virus to spread widely unseen.¹³ Vaccines that provide partial immunity and mask disease but allow hosts to continue to shed may speed viral evolution.⁹ Vaccination may have led to the evolution of more virulent H5N1 strains that evaded vaccine protection.

After AI hit Bangkok, a special hotline received nearly 1,200 reports of “mysterious” bird deaths. However, the birds, mostly budgerigars and parrots, were dying from shock and starvation after being released by their fearful owners. Most callers were so panicky they demanded an immediate diagnosis over the telephone. Up to 128,091 caged birds were waiting to be tested for AI. Bangkok crows were sampled after the deaths of two crows at a zoo were linked to an H5 AI. In Thailand 500 migratory open-billed storks and another 300 birds died in wetlands. Only 30–40% of the dead storks were infected with AI. There were no reports of AI in Bangladesh from where Asian open-bill storks migrate. Thus, the storks most likely were infected in Thailand.¹³

In late 2002, H5N1 killed non-domestic birds in parks and a zoo in Hong Kong, including waterfowl, greater flamingos, gray herons and egrets. In February 2003, avian H5N1 was isolated from two humans, one of whom died. Despite high genetic homology (>99.0% in all genes), the human isolates showed a very different reactivity pattern compared with the H5N1 viruses isolated from the wild waterfowl. All H5N1 isolates from 1997 to 2001 were non-pathogenic in ducks but the H5N1 isolates from late 2002 were highly pathogenic in ducks. This is the first time since 1961 that influenza viruses are known to have killed waterfowl.¹⁹ Despite the 2002 outbreak, there is little evidence the 2003/2004 H5N1 strain significantly affected wild bird populations or that wild birds spread it. Of 6000 wild birds tested in Hong Kong, one peregrine falcon was positive for the H5N1 strain. However, as the 2003/2004 H5N1 strain killed migratory wild birds, serologic studies in wild birds across Asia are needed to determine whether it became established in wild populations.⁴

Despite a lack of evidence, governments in China, Thailand, Cambodia, Japan and Hong Kong were quick to implicate wild birds in the spread of the virulent 2003/2004 H5N1.¹³ Responses ranged from the logical and effective to the fanciful and irresponsible. In China, authorities were required to monitor and disinfect the habitats of migratory birds, collect their excrement and sanitize it. Hong Kong closed parks and zoo exhibits but not poultry or wild bird markets. Despite the revelation that a chicken farm in Kyoto failed to report mass deaths due to H5N1 and continued to ship live chickens, eggs and meat while experiencing massive mortality, Japanese authorities supposed the virus was carried by migrating birds from Korea because there was no variation in the virus’ sequences in Japan.¹³

The Thai Agriculture Ministry wanted to cull migratory birds because killing almost 30 million chickens on 40,000 farms had not controlled the epizootic. In Thai provinces where AI re-emerged, the infection was found mostly in fighting roosters. Authorities suspected infected fighting roosters smuggled out of red zones during the first outbreak then returned to the areas had probably re-kindled the infection. Some owners refused to slaughter their prized fighting roosters. Yet a spokesperson for the University’s faculty of veterinary science said AI virus in yellow zones was due to the failure to eradicate all fowl and that birds in natural habitats should also be culled, not just chickens.¹³

Vaccination Trial

Virulent AI and SARS outbreaks in Hong Kong caused widespread fear and greatly reduced visitation to Ocean Park. Although H5N1 viruses are highly variable, there is much cross-protective immunity from H5N1 vaccines and non-pathogenic AI viruses such as H5N3.⁶ To allay public fear and to protect collection birds, we tested a killed H5N3 vaccine (HK/goose/1999 H5 and A/duck/Germany/1215/73 N3). This allows differentiation between infected and vaccinated birds by testing for different neuraminidase antibodies. The number of birds with protective titers (>1:16) 28 days after a single vaccination are shown in Table 1.

Table 1. Number of birds with protective titers (>1:16) 28 days after vaccination with experimental H5N3 vaccine

Vaccinations were repeated 28 days later (titers not yet available). Titers will also be determined 6 mo after the second vaccination to see whether protective levels last for the whole influenza season (6 mo). The titer levels indicate very high protective levels in most species. Three of 28 ducks and 3/7 swans maintained protective levels for over 12 mo following single prototype vaccinations in 2003. Results for psittacines are pending also but after the single 2003 prototype vaccination, 2/5 parrots developed protective levels.

Discussion and Recommendations

Live-bird markets provide outstanding conditions for genetic mixing and spreading of AI viruses²² and are critically important in the perpetuation and transmission of AI viruses to other avian species and to mammals, including humans¹⁵. In contrast, free-ranging wild birds appear to play a much lesser if any significant role in the ecology and epidemiology of virulent AI. Intensively reared poultry provide excellent opportunities for AI viruses to increase in virulence due to:

• Genetic uniformity in novel hosts providing intense advantage to mutant forms
• High density, large populations
• Confinement so that uninfected birds cannot avoid diseased birds or shed virus
• Food and water readily available to prolong the lives of severely ill birds
• Assisted transportation in vehicles to markets, farms and other businesses

Wild bird populations do not provide these opportunities due to:

• High genetic diversity avoiding intense selection for a single new mutant strain¹⁶
• Mostly low density, dispersed metapopulations
• Freely able to avoid ill birds and low risk of contacting contaminated sources
• Must work extremely hard for food under intense competition such that often even mildly affected individuals succumb rapidly, minimising pathogen spread
• No assistance for dispersal. Ill migratory birds unlikely to travel far.

Because all known influenza A subtypes exist in wild aquatic birds, avian influenza is not eradicable. Prevention and control are the only realistic goals.²² Governments should strive to:

• Monitor birds in live markets and exports/imports closely
• Close live bird markets while virulent AI is circulating in the region
• Improve biosecurity measures (e.g., prevent contact with wild aquatic spp.)
• Separate land-based poultry, pigs and aquatic avian species in farms and markets
• Monitor the movement of poultry between farms and markets closely
• Only allow controlled effective vaccination in response to virulent outbreaks
• Conduct serologic and other epidemiologic studies in wild birds across Asia to determine whether virulent AI became established in wild populations

Zoos and other organisations dealing with wild birds should:

• Avoid birds from commercial sources such as markets and dealers, especially those trading internationally
• Vaccinate at-risk or high-profile birds with an effective vaccine prior to the risk period
• Minimise the exposure of collection animals to wild birds, especially aquatic species
• Separate aquatic species from passerines, psittacines and other species

Literature Cited

1.  Alexander DJ, Parsons G, Manvell RJ. 1986. Experimental assessment of the pathogenicity of eight avian influenza A viruses of H5 subtype for chickens, turkeys, ducks and quail. Avian Pathol. 15: 647–662.

2.  Alexander DJ, Brown IH. 2000. Recent zoonoses caused by influenza A viruses. Rev Sci Tech. 19: 197–225.

3.  Capua I, Alexander DJ. 2002. Avian influenza and human health. Acta Trop. 83: 1–6.

4.  Dierauf L. Director. USGS National Wildlife Health Center.

5.  Guan Y, Peiris JSM, Lipatov AS, Ellis TM, Dyrting KC, Krauss S, Zhang LJ, Webster RG, Shortridge KF. 2002. Emergence of multiple genotypes of H5N1 avian influenza viruses in Hong Kong SAR. PNAS. 99: 8950–8955.

6.  Hiromoto Y, Yamazaki Y, Fukushima T, Saito T, Lindstrom SE, Omoe K, Nerome R, Lim W, Sugita S, Nerome K. 2000. Evolutionary characterization of the six internal genes of H5N1 human influenza A virus. J Gen Virol. 81: 1293–1303.

7.  Katz JM. 2003. The impact of avian influenza viruses on public health. Avian Dis. 47: 914–920.

8.  Kaverin NV, Rudneva IA, Ilyushina NA, Varich NL, Lipatov AS, Smirnov Y, Govorkova EA, Gitelman AK, Lvov DK, Webster RG. 2002. Structure of antigenic sites on the haemagglutinin molecule of H5 avian influenza virus and phenotypic variation of escape mutants. J Gen Virology. 83: 2497–2505.

9.  Liu M, Wood JM, Ellis T, Krauss S, Seiler P, Johnson C, Hoffmann E, Humberd J, Hulse D, Zhang Y, Webster RG, Perez DR. 2003. Preparation of a standardized, efficacious agricultural H5N3 vaccine by reverse genetics. Virology. 314: 580–590.

10.  Mase M, Imada T, Sanada Y, Etoh M, Sanada N, Tsukamoto K, Kawaoka Y, Yamaguchi S. 2001. Imported parakeets harbor H9N2 influenza A viruses that are genetically closely related to those transmitted to humans in Hong Kong. J Virol. 75: 3490–3494.

11.  Nestorowicz A, Kawaoka Y, Bean WJ, Webster RG. 1987. Molecular analysis of the hemagglutinin genes of Australian H7N7 influenza viruses: role of passerine birds in maintenance or transmission? Virology. 160: 411–418.

12.  Perkins LEI, Swayne DE. 2003. Varied pathogenicity of a Hong Kong–origin H5N1 avian influenza virus in four passerine species and budgerigars. Vet Path. 40: 14–24.

13.  ProMED-mail. International Society for Infectious Diseases. https://promedmail.org/.

14.  Sharp GB, Kawaoka Y, Wright SM, Turner B, Hinshaw V, Webster RG. 1993. Wild ducks are the reservoir for only a limited number of influenza A subtypes. Epidemiol Infect. 110: 161–176.

15.  Shortridge KF, Zhou NN, Guan Y, Gao P, Ito T, Kawaoka Y, Kodihalli S, Krauss S, Markwell D, Murti G, Norwood M, Senne D, Sims L, Takada A, Webster RG. 1998. Characterization of avian H5N1 influenza viruses from poultry in Hong Kong. Virology. 252: 331–342.

16.  Spielman D, Brook BW, Briscoe DA, Frankham R. 2004. Does inbreeding and loss of genetic diversity decrease disease resistance? Cons Biology. In print.

17.  Sturm-Ramireza KM, Ellis T, Bousfield B, Guand Y, Peiris M, Webster R. 2004. H5N1 Influenza A viruses from 2002 are highly pathogenic in waterfowl. J Virology. In press.

18.  Suarez DL. 2000. Evolution of avian influenza viruses. Vet Microbiol. 74: 15–27.

19.  Suzuki Y, Nei M. 2002. Origin and evolution of influenza virus hemagglutinin genes. Mol Biol Evol. 19: 501–509.

20.  Webby RJ. Influenza virologist. St Jude’s Children’s Research Hospital. Memphis, Tennessee, USA.

21.  Webby RJ, Webster RG. 2001. Emergence of influenza A viruses. Philos Trans R Soc Lond B Biol Sci. 356: 1817–1828.

22.  Webster RG. 1998. Influenza: an emerging disease. Emerg Infect Dis. 4: 436–441.

23.  Webster RG, Sharp GB, Claas EC. 1995. Interspecies transmission of influenza viruses. Am J Respir Crit Care Med. 152: S25–S30.

24.  Zhou NN, Shortridge KF, Claas ECJ, Krauss SL, Webster RG. 1999. Rapid evolution of H5N1 influenza viruses in chickens in Hong Kong. J Virol. 73: 3366–3374.