IFST Information Statement on Avian Influenza and Food

The highly pathogenic H5NI strains of avian flu which are transmitted from wild birds to domestic poultry are currently transmissible to humans only with great difficulty, as a result of close direct contact with live or dead infected birds. This is probably because the receptors for H5N1 in humans are deep in the lungs but not in the nose, throat or upper respiratory tract.

When H5N1 is present in poultry, the virus can be present in meat and eggs from affected birds. However, if poultry meat is handled with good hygiene and meat or eggs are properly cooked (the same precautions needed to avoid bacterial food poisoning) there is no risk to humans consuming them.

The worst case scenario for human health is if H5N1 mutates into a form which, while retaining its virulence, could attach to receptors in the nose, throat or upper respiratory tract. The mutated virus could then be more easily caught by humans and transmitted from human to human, leading to a global pandemic. Such an event could theoretically occur by genetic recombination if a person suffering from flu caused by a 'normal' human strain of influenza virus acquired an H5N1 virus, which itself had little capacity to infect humans. The World Health Organisation, UN Food and Agriculture Organisation, EU Commission and many national governments are making plans to try to avert or cope with such an eventuality.

Following poultry outbreaks of H5N1 avian flu in many countries, an outbreak of H5N1 at a turkey farm in the UK in early February 2007 has proved to be essentially identical to the strain that occurred in two flocks of geese in Hungary in late January 2007, and investigations have concluded that the most plausible vector was imported semi-processed raw turkey meat from Hungary; that there was no evidence that any meat entered the UK food chain from the restricted zones in Hungary; and that the risk to the health of the workers in the processing plant, or the wider poultry farm was very low.

Avian Influenza

To bird experts, avian flu is nothing new; it first appeared in Italy around 1878. There are at least 144 types of bird flu viruses, most of which do not kill the birds.

The influenza viruses are single stranded RNA viruses with segmented genomes, of the Orthomyxoviridae family. They can be classified into A, B and C types. All these infect humans, but only influenza A viruses are found in birds and other animals. A wide range of influenza A viruses has been described. They are subtyped based on sequencing or antigenic characterisation of the two biologically important surface proteins; the haemagglutinin (H) and the neuraminidase (N). Fifteen H types and 9 N types have been described, and they are found in different combinations. All H and N type subtypes occur in birds, indicating their role as the main reservoir, but only a limited range of subtypes have been shown to circulate in humans (currently H3N2, H1N1). Avian influenza strains are commonly described as either ‘low pathogenicity’ or ‘high pathogenicity’ strains based on differences in the haemagglutinin gene and disease caused in chickens. One important biological feature of influenza viruses influencing their infectivity is the ability to bind to cells to initiate infection. This receptor binding specificity varies amongst influenza strains; avian influenza strains preferentially bind to sialic acid a 2.3-galactose linkage, human strains to a 2.6- galactose. Both receptors are found in the upper respiratory tract epithelia of pigs, but only the a 2.3 receptors are found in birds and the majority of receptors on human tracheal epithelial cells are of a 2.6 specificity. Avian influenza viruses are enveloped RNA viruses. AI strains appear to be more resistant to lower pH (3 months at 4°C, but much higher rates of inactivation occur at higher temperatures (>6 x 106 loss of infectivity within 2 weeks at 20°C). (ACMSF, 2005).

In Asia, the H5N1 virus went from wild birds to domestic poultry, evolved in poultry and re-infected wild birds becoming more lethal in the process. Indeed, the process of low pathogenic avian influenza (LPAI) evolving to become highly pathogenic avian influenza (HPAI) is well recognised in scientific literature. It kills some wild birds, but not all. In fact wild birds can be carriers of the virus and not become ill. Poultry flocks, ducks and geese have been infected, and this has given rise to obvious concerns. It has been argued that national and organisational strategies should be developed, that ensure an integrated, rapid response to foreign animal disease incidents in order to minimise the human, animal welfare and economic impact of such an event (WHO, 2002; Manning et al., 2005).

Transmission to humans?

H5N1 is transmissible to humans but currently only with difficulty. Hundreds of millions of birds have died of H5N1 infection since 2003. Yet the most up-to-date and comprehensive cumulative total of reported human cases maintained by the World Health Organisation, as at 19 February 2007, shows 274 reported human cases (mostly in Asia), of whom 167 died. There has been a disproportionate concentration of infections in children and young adults, even allowing for the relatively young populations in the eleven countries where human infections have occurred, and there is an over-representation of females among patients aged 10-29 years.

A speculative suggestion has been made of immunity over the ago of 40. However, the data are thought by WHO to be related to the fact that it is usually young people and women who look after domestic poultry. Current evidence indicates that most human infection is associated with close direct contact with live or dead infected birds. However, a minority of cases of human infection in Vietnam could not be related to close direct contact and it has been speculated that it may be related to consumption of uncooked infected poultry or of water contaminated with infected bird or poultry faeces.

Only recently research (Shinya et al, 2006) has indicated a probable reason why it is currently so difficult for humans to catch H5N1. Receptors for H5N1 are deep in the lungs but not in the nose, throat or upper respiratory tract. This is consistent with the risk assessment by ACMSF referred to above. However, If the virus mutates (evolves), to the point where it could pick up the ability to infect the nose, throat etc it would thereby easily be caught by humans and spread from human-to-human. If it managed to do this and to hold on to its virulence (potency, ability to kill), we could be facing a serious global flu pandemic.

The H7N7 AI virus that caused an outbreak on 225 poultry farms in Holland was associated with conjunctivitis in humans (Abbott, 2003). There was also very limited human-to-human transmission of the virus and evidence of infection in pigs. (van Kolfschooten, 2003). Fouchier et al., (2004) stated that the highly pathogenic AI virus was closely related to low pathogenic isolates from wild ducks. The same virus was detected in 86 humans who were in contact with poultry and three of their family members; of these 78 presented with conjunctivitis, 5 with conjunctivitis and flu-symptoms, 2 with flu symptoms and 4 did not show these symptoms. There was one fatal case of pneumonia with acute respiratory problems. The virus isolated from the fatal case displayed 14 amino acid substitutions which Fouchier et al. (2004) argued could be associated with enhanced disease and demonstrated the potential for a pandemic threat to humans.

Yamata et al (2006) point out that a conversion from sialic acid a 2.3-galactose to sialic acid a 2.3-galactose recognition is thought to be one of the changes that must occur before avian influenza viruses can replicate efficiently in humans and acquire the potential to cause a pandemic. By identifying mutations in the receptor-binding haemagglutinin (HA) molecule that would enable avian H5N1 viruses to recognize human-type host cell receptors, it may be possible to predict (and thus to increase preparedness for) the emergence of pandemic viruses. They show that some H5N1 viruses isolated from humans can bind to both human and avian receptors, in contrast to those isolated from chickens and ducks, which recognize the avian receptors exclusively. Mutations at positions 182 and 192 independently convert the HAs of H5N1 viruses known to recognize the avian receptor to ones that recognize the human receptor. Analysis of the crystal structure of the HA from an H5N1 virus used in these genetic experiments shows that the locations of these amino acids in the HA molecule are compatible with an effect on receptor binding. These changes might serve as molecular markers for assessing the pandemic potential of H5N1 field isolates.

Possibility of a future flu pandemic

If avian flu is contracted by a person already suffering from a human flu strain, there is the possibility of gene exchange and that the H5N1 virus will mutate. If this occurs there is the possibility that it would mutate into a virulent strain highly infectious to humans and transmissible between humans. What is going to happen or when are unknowns. What is certain is that the H5N1 strain has already undergone many changes. The strains found Hungary are different in many respects from those found initially in the far east. However, fortunately the changed strains still do not readily infect humans. We should, however, be aware of the fact that if/when such a change has occurred that will not be the end of the matter — more changes may well continue to occur presumably ad infinitum just as they do now with existing strains of human A influenza virus. How important such changes will be, is unknown and there are many possibilities, one of which is that at some point there will develop a virulent strain highly infectious to humans and transmissible between humans.

This could give rise to a global pandemic in which millions would become ill and many would die. Three human influenza pandemics have occurred during the 20th century (Manning et al., 2005). The 1918-1919 pandemic led to an estimated 20 – 40 million deaths worldwide (Reid et al., 1999). There was a mild wave of influenza in the spring and summer of 1918 that was highly contagious but caused few human deaths. In late August, a more virulent form emerged. This mirrors the behaviour of the avian influenza virus in bird populations with low pathogenic and highly pathogenic forms. Both the 1957 and 1968 human influenza pandemics arose in southern China, a region regarded as an epicentre for the emergence of pandemic influenza (Shortridge and Stuart-Harris (1982) cited by Guan et al., 2002).

In a recent review (Thomas and Noppenberger, 2006) conclude that avian influenza virus A/H5N1 is a public health threat that has the potential to cause serious illness and death in humans. Understanding its pathology, transmission, clinical features, and pharmacologic treatments and preparing for the prevention and management of its outbreak will help avoid its potentially devastating consequences.

Strategies to prevent and treat a flu pandemic

Among the catastrophic societal consequences of a modern flu pandemic would be the difficulty of maintaining production and distribution of the food supply. Governments and food companies need to plan ahead for this “worst case” scenario.

Faced with the threat of an influenza pandemic, governments around the world are developing strategies to prevent and treat a pandemic. In March 2006 The Royal Society and the Academy of Medical Sciences launched a joint study to examine the extent to which scientific evidence is being incorporated into preparedness for a pandemic, and to identify areas where policy makers should make better use of the scientific evidence in policy development and contingency planning (Royal Society, 2006).

The World Health Organisation has issued a “pandemic influenza draft protocol for rapid response and containment” (WHO, 2007). It points out that containment of a potential pandemic has never been attempted; but the world has never before received an advance warning that a pandemic may be imminent. The practical and logistics challenges are formidable and success is not assured. Nonetheless, the strategy should be pursued for several compelling reasons – successful containment will avert an enormous amount of human suffering and possibly millions of deaths, while also sparing the world considerable economic and social disruption.

“Even if containment efforts ultimately fail to stop the emergence of a fully fit pandemic virus, these efforts could slow the initial spread of the pandemic and give countries time to increase preparedness. Each day gained following the emergence of a pandemic virus – if rapidly detected – allows the production of around 5 million doses of a pandemic vaccine. Each added day gives countries more time to adapt routine health services to an emergency situation. Time gained also allows WHO to predict patterns of further spread and issue appropriate alerts.” (WHO, 2006a, 2007).

Likewise the European Commission has published guidance on preparedness (EU Commission, 2006a, 2006b); and the US has issued an Implementation Plan for the National Strategy for Pandemic Influenza. (Homeland Security Council, 2006; USDA, 2006).

How do you plan for a scenario where, estimates suggest, anything from 25% to 40% of the workforce would be incapacitated and possibly many more staying home from fear of infection?

Of course you cannot isolate the food supply from other factors that would be equally affected – like transportation and public utilities — but what measures could food manufacturers and retailers take? There are two areas in which the food supply would be most vulnerable – transfer of infection to and among employees, and the “just-in-time” philosophy embraced by the food industry. It is suggested these are the areas for attention to minimise the impact of those vulnerabilities. Employees actually handling the food cannot work from home. Obvious steps common to all employers should be taken, such as providing employees with face masks and enforcing strict rules on hygiene, especially hand-washing, both in factory operations and in canteens. Arrangements could be planned for other employees to work from home by computer. Face-to-face meetings could be replaced by teleconferences or e-mailing on listservs or via websites.

“Just-in-time” methods would have to be abandoned in favour of building up inventories of raw materials, packaging materials and manufactured products. This would have to be planned and organised by close collaboration among suppliers, manufacturers and retailers. Building up inventories would mean that extra storage space would have to be found.

One major UK retailing group, Sainsbury, has said that it plans to cater for higher demand for the types of products people buy when ill, such as painkillers, soups and “store-cupboard foods” (Report in The Grocer, 4 March 2006).

Meat and eggs from infected birds

Meanwhile, it is important to prevent, or minimise the risk of, exposure of people to infected poultry or poultry products.

In countries where avian flu is present in poultry, the virus can be present in meat and eggs from affected birds. That is why some countries have imposed bans on imports of poultry and poultry products from counties where flocks have been infected. As time goes by and inevitably infection events occur in more and more countries, this precaution, while remaining effective in some circumstances, overall will become less and less effective as a mechanism to control the spread of avian flu.

The greatest risk of exposure to the virus is through the handling and slaughter of live infected poultry. Good hygiene practices are essential during slaughter and post-slaughter handling.

In catering and in the home, good hygiene practices are necessary to prevent exposure via raw poultry meat or cross contamination from poultry to other foods, food preparation surfaces or equipment or vehicles.

When handling raw poultry, the person involved in the food preparation should wash their hands thoroughly — soap and warm water are sufficient for this purpose – and dry them thoroughly They should clean surfaces and utensils that have come into contact with the poultry products. Thus, exactly the same precautions are needed as those to prevent food poisoning by other pathogens. Likewise precautions must be taken to avoid recontamination of cooked products.

Even if the virus were present in meat or eggs, several factors will contribute to preventing or limiting its effects on people. Firstly, the. H5N1 strain of avian flu is sensitive to heat and is killed at normal temperatures used for cooking (70°C in all parts of the food). In fact, cooking infected chicken meat adequately to kill other pathogens provides a large safety margin in respect of the killing the virus (Thomas and Swayne, 2006). Secondly, even if it is still present after cooking, the virus is destroyed by saliva and by gastric acid, as well as the fact that there are in the human gut very few of the receptors the virus would need to enter the body.

In relation to eggs from infected birds, WHO states that both the white and yolk must be firm when cooked and that this advice is precautionary for all bacteria and viruses that may be present, for all parts of the world. FSA originally stated, in “Questions and Answers” on its website, that only the white need be firm; that it is not necessary to cook eggs until the yolks are hard to protect against exposure to the avian flu virus. However, eggs from H5N1 infected birds cannot be assumed to be free from other pathogens. FSA Chief Scientist Andrew Wadge has confirmed that any FSA advice on avian influenza will be cross-referenced to FSA’s long standing advice to take especial care when catering for the elderly, the sick, babies and toddlers, and pregnant women. These groups should not be served raw or partially cooked (e.g. soft-boiled) eggs or products.

It is important to avoid access to potentially infected poultry or poultry products, including discarded scraps, by cats or rodents or scavenging birds such as seagulls. FAO has advised that cats can become infected with the H5N1 avian influenza virus, but at present there is no scientific evidence to suggest that there has been sustained transmission of the virus in cats or from cats to humans (FAO, 2007). As a precautionary measure, FAO recommended that in areas where the H5N1 virus has been found in poultry or wild birds, cats should be separated from infected birds until the danger has passed. On commercial poultry premises cats should even be kept indoors.

Meat or eggs from vaccinated poultry

Should vaccination of poultry be carried out? Veterinarians recognise that there are pros and cons. However two further questions – whether it is safe to eat meat and eggs from vaccinated birds and, if so, whether the public would accept assurances from scientists and/or the authorities to this effect. The vaccines used to vaccinate birds against avian flu do not pose any human health concerns — provided that a licensed vaccine with marketing authorisation is used, and the correct interval is observed between vaccination and slaughter or the date the eggs are laid. However, there is concern that vaccination of poultry may mask infection in the vaccinated birds making them potentially infectious, but asymptomatic.

Avian influenza outbreak in the UK

Following poultry outbreaks of H5N1 avian flu in many countries (WHO, 2007b), on 2 February 2007, the European Commission was informed by the UK authorities of a suspected outbreak of avian influenza in Suffolk, in the east of England (Defra, 2007a). The outbreak occurred on a Bernard Matthews farm of 159 000 turkeys, and was detected following the death of around 2500 birds. The Community Reference Laboratory in Weybridge swiftly confirmed the disease to be the H5N1 strain of avian influenza. The UK authorities were already applying the measures laid down in the Avian Influenza Directive and Decision 2006/415/EC on avian influenza in domestic poultry; the establishment of a protection zone (PZ) of 3 km radius and a surveillance zone (SZ) of 10 km around the infected holding. The area is classified as a high risk area (area A) which is surrounded by a low risk area (area B) acting as a buffer zone to the disease free parts of the country. This involved strict movement controls, poultry must be kept indoors, a prohibition on gatherings of poultry and other birds and strengthening of on-farm biosecurity measures.

On 3 February, In consultation with ornithologists an additional wider Restricted Zone (RZ) was imposed covering approximately 2090sqkm requiring the isolation of poultry from wild birds and movements to be licensed. The national general licence on bird gatherings was revoked, and bird shows and pigeon racing not permitted.

Although the dead birds were from one shed out of 20, all 159,000 birds were culled. Subsequently it was found that birds in three other sheds had been infected. Further investigations showed that the virus strain was essentially identical to the virus strain found on 29 January 2007 in two separate flocks of geese in Hungary. It then transpired that Bernard Matthews has been importing partly-processed raw turkey meat from a plant in Hungary to a company processing plant close to the outbreak farm. This gave rise to a further investigation into whether the vector for the UK outbreak could be poultry-to-poultry rather than wild bird-to-poultry.

It also gave rise to a Food Standards Agency investigation. FSA stated:
“Our advice, that avian flu does not pose a food safety risk, remains unchanged. However, it is illegal for infected meat to be in our food and so the Agency would take any appropriate action if it were found to be there.” (FSA, 2007a).

On 16 February the results of the investigations by FSA, Defra , the Health Protection Agency (HPA)and the Meat Hygiene Service (MHS) were published (FSA, 2007b)

The Defra interim epidemiological report identifies two possible hypotheses for the introduction of H5N1 into the poultry premises at Suffolk:

·the report concludes that there is 'little evidence' to support the first hypothesis of transmission from a wild bird source. This draws on advice from expert ornithologists and the fact that H5N1 has not been found in the wild bird population in Europe since August 2006. In addition to this, extensive surveillance from the Infected Premises and the surrounding area has not isolated any trace of H5N1 in wild birds

·the second hypothesis examined in the report is the spread of the virus associated with the importation of poultry products from Hungary. This is supported by the final virology results from the Veterinary Laboratories Agency (VLA) confirming that the virus strain found in poultry in Suffolk is 'essentially identical' to that which caused the outbreaks in Hungary

The interim report therefore concludes that 'currently the most plausible' route of transmission is associated with the importation of poultry products via Hungary.

The joint final report by the FSA, Defra, HPA and MHS examined transmission via imported Hungarian turkey meat. The FSA-led part of the investigation was launched to check whether meat from a restricted zone in Hungary had been brought to the Bernard Matthews plant at Holton, Suffolk.

This followed the hypothesis that there may be a link between the Hungarian outbreaks and the avian influenza outbreak in Suffolk. If it had been discovered that meat exported from Hungary to the UK had come from inside an avian influenza restricted zone, this would have been illegal under EU law.

Its main findings are:

·there is no evidence that any meat entered the UK food chain from the restricted zones in Hungary

·from evidence gathered by the FSA investigation team it appears that all food importing and processing activities being undertaken at the Bernard Matthews factory at Holton are in line with EC law

·the outbreak of H5N1 avian influenza does not alter the FSA's advice that properly cooked poultry meat remains safe to eat

The HPA’s investigation focused on establishing if there was any health threat to the workers