30 Influenza

Influenza is a highly infectious, acute illness which has affected humans and animals since ancient times. Influenza viruses form the Orthomyxoviridae family and are grouped into types A, B, and C on the basis of the antigenic nature of the internal nucleocapsid or the matrix protein. Influenza A viruses infect a large variety of animal species, including humans, pigs, horses, sea mammals, and birds, occasionally producing devastating pandemics in humans, such as in 1918 when it has been estimated that between 50–100 million deaths occurred worldwide.

There are two important viral surface glycoproteins, the haemagglutinin (HA) and neuraminidase (NA). The HA binds to sialic acid receptors on the membrane of host cells and is the primary antigen against which a host’s antibody response is targeted. The NA cleaves the sialic acid bond attaching new viral particles to the cell membrane of host cells allowing their release. The NA is also the target of the neuraminidase inhibitor class of antiviral agents that include oseltamivir and zanamivir and newer agents such as peramivir. Both these glycoproteins are important antigens for inducing protective immunity in the host and therefore show the greatest variation.

Influenza A viruses are classified into 16 antigenically distinct HA (H1–16) and 9 NA subtypes (N1–9). Although viruses of relatively few subtype combinations have been isolated from mammalian species, all subtypes, in most combinations, have been isolated from birds. Each virus possesses one HA and one NA subtype (Tables 30.1 and 30.2).

Last century, the sudden emergence of antigenically different strains in humans, termed antigenic shift, occurred on three occasions, 1918 (H1N1), 1957 (H2N2) and 1968 (H3N2), resulting in pandemics. The frequent epidemics that occur between the pandemics are as a result of gradual antigenic change in the prevalent virus, termed antigenic drift. Epidemics throughout the world occur in the human population due to infection with influenza A viruses, such as H1N1 and H3N2 subtypes, or with influenza B virus. Phylogenetic studies have led to the suggestion that aquatic birds that show no signs of disease could be the source of many influenza A viruses in other species. The 1918 H1N1 pandemic strain is thought to have arisen as a result of spontaneous mutations within an avian H1N1 virus. However, most pandemic strains, such as the 1957 H2N2, 1968 H3N2 and 2009 pandemic H1N1, are considered to have emerged by genetic re-assortment of the segmented RNA genome of the virus, with the avian and human influenza A viruses infecting the same host.

Influenza viruses do not pass readily between humans and birds but transmission between humans and other animals has been demonstrated. This has led to the suggestion that the proposed re-assortment of human and avian influenza viruses takes place in an intermediate animal with subsequent infection of the human population. Pigs have been considered the leading contender for the role of intermediary because they may serve as hosts for productive infections of both avian and human viruses, and there is good evidence that they have been involved in interspecies transmission of influenza viruses; particularly the spread of H1N1 viruses to humans. Apart from public health measures related to the rapid identification of cases and isolation, the main control measures for influenza virus infection in human populations involves immunization and antiviral prophylaxis or treatment.

The highly infectious, acute respiratory illness now known as influenza has affected human beings since ancient times. The individual symptoms and epidemiological characteristics of the disease are sufficiently distinct that it is possible to identify a number of major epidemics in the distant past. One such epidemic was recorded by Hippocrates in 412 bc, and numerous episodes were described in the Middle Ages.

The name influenza has its origins in early fifteenth century Italy and was adopted in Europe to explain the sudden and unexpected appearance of an epidemic disease thought to be under the influence of the stars (Kaplan and Webster 1977).

The first well-recorded pandemic in humans, in which mortality was frequently high, particularly in densely populated areas, occurred in 1580 and was believed to have originated in Asia before spreading to Africa and Europe. During the following three centuries, although record keeping was irregular and reporting was often inaccurate, there were recognizable clinical accounts of a number of serious influenza pandemics. Retrospective research partially identified the virus responsible for the first pandemic of 1889 by testing for influenza antibodies in serum of people who were alive at that time (Tumova 1980).

Possibly the most devastating influenza pandemic recorded occurred in 1918. It has been estimated that during this pandemic between 50–100 million deaths occurred throughout the world and that in resource rich countries, such as the USA, about 0.5% of the population died. In some parts of Alaska and the Pacific islands more than half the population was lost. In the USA alone it is estimated that life expectancy following the 1918 pandemic dropped by 12 years. There was an enormous impact on society in terms of mortality, morbidity, and economic factors. At the height of the epidemic community life in many cities was brought almost to a standstill. The repercussions of the pandemic were felt by armed forces engaged in the First World War, with some 43, 000 deaths in the USA forces alone, representing about 80% of the total number of USA battle deaths in the war. The 1918 to 1919 influenza pandemic was caused by an influenza A H1N1 virus and sequence analysis of virus recovered from autopsy paraffin-embedded formalin fixed lung tissue suggested an avian source that became adapted to mammals (Reid et al. 2000). The virus is genetically distinct from any of the avian or mammalian influenza virus sequences analysed from that time onwards (Basler et al. 2001; Reid et al. 2003, 2004; Taubenberger et al. 1997, 2003, 2005). Clinical and autopsy reports have demonstrated that the high morbidity and mortality rates were due to aggressive brochopneumonia due to secondary bacterial infections (Morens et al. 2010). At that time, the range of antibiotics that can now be used to effectively treat these infections were not available. In addition, there may have been an immunopathogenic cause leading to acute respiratory distress syndrome (ARDS) due to cytokine storms in some individuals. There was a high mortality rate at all ages, but especially among the 20 to 40 year age group. This may have been due to a vigorous immune response characterized by a release of pro-inflammatory cytokines or an antibody dependent enhancement of the response having been exposed to other influenza viruses in the past (Cheung et al. 2002).

For centuries there had been wild speculation regarding the cause of influenza, but by the end of the nineteenth century the microbiological concept of infectious disease had become accepted. Following on from this was the discovery of a bacillus in the throats of many influenza patients. This bacillus, Haemophilus influenzae, was for many years the leading suspect for the causative agent of influenza. The first evidence of the true viral cause came in the late 1920s when a virus was found in pigs showing disease similar to influenza in humans and successfully transmitted between pigs using filtered material. A related strain was finally isolated from a human patient in 1933 by inoculating a filtrate of throat washings into the noses of ferrets. Rapid progress followed the demonstration that influenza virus could be transmitted to ferrets and mice.

A second type of influenza virus from man was transmitted experimentally to ferrets in 1940. This virus was designated influenza B to distinguish it from the first type found, which became known as influenza A. In 1933 it was found that influenza viruses would multiply in cells lining the allantoic cavity of the developing chicken embryo. This was followed by the observation that infective allantoic fluid caused agglutination of chicken red blood cells, also referred to as haemagglutination. These developments laid the foundations for early work on influenza viruses and these techniques are still used when carrying out work with influenza viruses. The haemagglutination reaction could be inhibited by specific serum antibodies to influenza viruses, thereby a simple assay could facilitate strain differentiation and the detection of an individual’s immunological response to an influenza virus infection.

The history of influenza in animals is equally confused, not least because influenza may be used as a general term for respiratory illness in animals, as it is in humans, and the wide range of infectious agents that can infect the upper respiratory tract causing influenza-like signs. However, there are many historical reports of influenza occurring simultaneously or sequentially in humans and domestic animals (Beveridge 1977) suggesting an early understanding of a possible link between the disease in animals and humans. Close correlation between human and animal influenza was made during the 1918 pandemic, and the term swine influenza was applied to a ‘new’ disease of pigs described at that time which produced clinical signs similar to those in humans (Dorset et al. 1922). Following the first isolation of influenza virus it was known that H1N1 (Hsw1N1) virus remained endemic in pig populations, particularly in the USA.

Little consideration was given to the possibility of influenza infections of other animals until 1955. In that year it was demonstrated that the causative virus of a highly pathogenic disease of chickens known as ‘fowl plague’, which had been isolated and described as a filterable agent as early as 1901 (Centanni and Savonuzzi, cited by Stubbs 1965), was a type A influenza virus. Several, less virulent, viruses that had been isolated from domestic poultry up to that time were also shown to be influenza A viruses. In 1956, evidence was obtained of influenza A virus infections in horses (Sovinova et al. 1958) and in the next two years respiratory disease in horses caused by this virus became widespread in Europe. These findings aroused the interest of many scientists working on influenza in humans, which was further concentrated by the H2N2 ‘Asian flu’ pandemic of 1957. Since that time the World Health Organization (WHO) and World Organization for Animal Health (OIE) have coordinated work on the epidemiology of animal viruses, particularly in relation to human influenza. However, it was not until the late 1970s that the true picture of the vast reservoir of influenza viruses that exist in animals, particularly birds, had been formed.

Avian influenza (AI) infections became a major focus after bird to human transmission occurred in Hong Kong in 1997, resulting in severe infections and a number of deaths in humans. Influenza A virus infections are endemic in pigs, horses, ducks, geese, swans, shorebirds and domestic poultry. Sporadic infections have been seen in farmed mink, whales, seals, dogs, and tigers and leopards in captivity. In Europe and Asia there have been reports of small outbreaks of AI virus infections involving H5N1, H7N7 and H9N2 viruses (Keawcharoen et al. 2004).

In April 2009, there were reports from Mexico and Southern California of a respiratory illness caused by a novel swine influenza H1N1 virus. By June 2009, WHO had declared pandemic phase 6 as pandemic H1N1 was reported globally. This virus contained gene re-assortments from Eurasian and North American swine influenza, North American avian influenza and North American human influenza virus infections (Neumann et al. 2009; Dawood et al. 2009; Zimmer et al. 2009).

The influenza viruses belong to the Orthomyxoviridae family of which there are five genera, influenza type A, B and C viruses, Thogoto viruses and Infectious Salmon Anaemia Virus (International Committee on Taxonomy of Viruses 2009 website http://www.ICTVdb.org/Ictv/index.htm).

Influenza viruses are grouped into types A, B, and C on the basis of the antigenic nature of the internal nucleocapsid or the matrix protein. Both these antigens are common to all viruses of the same type. Viruses of influenza A type are further divided into subtypes on the basis of the haemagglutinin (HA) and neuraminidase (NA) antigens. As of 2010, 16 HA and nine NA subtypes have been recognized; each virus possesses one HA and one NA subtype.

Influenza virus types A, B, and C infect humans, but, except for occasional reports, infections of other animals are restricted to influenza A viruses. Only influenza A viruses have been isolated from birds. Types A and B viruses both cause similar clinical disease in humans and both may be responsible for epidemics in humans. However, only influenza A viruses have produced the devastating pandemics that have made such an impact on the human population throughout recorded history.

Influenza A virus particles appear roughly spherical or filamentous, 80–120 nm in diameter. The nucleocapsid shows helical symmetry and is enclosed within a protein matrix. External to the matrix is a lipid membrane, the surface of which is covered by two types of glycoprotein projections, or spikes, with which haemagglutinin and neuraminidase activities are associated. These two surface glycopeptides, particularly the haemagglutinin, are the most important antigens stimulating protective immunity in the host. Consequently, considerable antigenic variation is seen in these polypeptides while other polypeptides are antigenically more stable.

The genomes of influenza A and B viruses consist of eight unique segments of single-stranded RNA which are of negative polarity. Replication and transcription occur in the host cell nucleus. Influenza C viruses possess seven segments of RNA. The viral RNA is transcribed to complementary messenger RNA by a virus-associated polymerase complex (designated PB1, PB2, and PA). To be infectious, a single virus particle must contain each of the eight unique RNA segments. It is likely that the incorporation of RNAs into the virion is at least partly random. The random incorporation of RNA segments allows the generation of progeny viruses containing novel combinations of genes when cells are infected with two different parent viruses. This phenomenon is referred to as genetic re-assortment.

The eight influenza A viral RNA segments encode 11 proteins that include PB1, PB2, and PA polymerases, HA, nucleoprotein (NP), NA, matrix proteins (M1 and M2), and non-structural proteins (NS1 and NS2).

The three largest proteins (PB1, PB2 and PA) and one intermediate size protein (NP) are found in the RNA polymerase complex, which has transcriptase and endonuclease activities. This complex is involved in the synthesis of the three classes of virus-specific RNA molecules detected in infected cells, mRNA, virion RNA (vRNA), and complementary RNA (cRNA). PB2 functions during the initiation of viral mRNA transcription, recognizing the 5′ terminal caps of host cell mRNAs for use as viral mRNA transcription primers and is involved in the endonucleolytic cleavage of these primers. PB1 is responsible for the elongation of the primed nascent viral mRNA, template RNA and vRNA. The PA has a number of roles in replication. NP is transported into the infected cell nucleus, where it binds to and encapsidates viral RNA. NP is phosphorylated, the pattern of which is host cell dependent and may be related to viral host range restriction. NS1 has a role in virulence as it is an interferon antagonist that blocks the activation of transcription factors. It also binds to dsRNA and prevents RNase L activation thereby affecting the innate immune response to influenza virus infection. The NS1 proteins have also been associated with high levels of pro-inflammatory cytokines in the host, that may lead to more morbidity and mortality.

Influenza viruses enter the host cell via the sialic acid receptor. Endosomal entry and acidification activates the viral M2 ion channel and uncoating of the virus occurs as a result due to the disruption of the M1-viral ribonucleoprotein (RNP) complex. The HA-2 component is conformationally rearranged as a result of the pH change and leads to endosomal membrane fusion. Single vRNPs are released into the cytoplasm.

Nuclear localization signals carried by the NP protein and the importin alpha–beta mediated pathway assist the negative sense vRNA to be transported into the nucleus. These vRNAs are vRNP complexes associating heterotrimeric RNA dependent RNA polymerase (RdRp, made up of PB1, PB2, PA proteins) and nucleoprotein (NP).

An initial round of transcription produces 5′ capped and 3′ poly(A) viral mRNA exported towards the cytoplasm to be translated. Then vRNA templates are replicated into a positive sense copy RNA (cRNA).

Termination occurs by generating a poly(A) tail. After the formation of vRNPs within the nucleus, M1, NEP (Nuclear Export Protein, primary referred to as NS2) and NP catalyse its transport to the cytoplasm. M1-vRNP complexes are directed to the assembly site where both HA and NA have been accumulated. M1 interacts with cytoplasmic tails of glycoproteins, leading to assembly and budding of virions. The viral NA sialidase acts to release virions from the cell surface (Jossetab et al. 2008).

The HA protein is an integral membrane protein and the major surface antigen of the influenza virus virion. It is responsible for the binding of virions to host cell receptors and for fusion between the virion envelope and the host cell. Newly synthesized HA is cleaved to remove the amino-terminal hydrophobic sequence which is the signal sequence for transport to the cell membrane. Carbohydrate side-chains are added, the number and position of which vary with the virus strain. Palmitic acid is added to cysteine residues near the HA carboxy terminus. Fusion activity requires post-translational cleavage of HA by cellular proteases into the disulphide-linked fragments HA1 and HA2. This cleavage of the HA does not affect its antigenic or receptor-binding properties, but is essential for the virus to be infectious and is an important determinant in pathogenicity. HA molecules form homotrimers during maturation. The three-dimensional structure of the complete trimer has been determined, consisting of a globular head (HA1) on a stalk (HA1/2). The head contains the receptor-binding cavity as well as most of the antigenic sites of the molecule. The carboxy terminus of HA2 anchors the glycoprotein in the cell or virion membrane. The HA is subject to a high rate of mutation due to error-prone viral RNA polymerase activity. Selection for amino acid substitutions is driven at least in part by immune pressure, as the HA is the major target of the host immune response. The amino acids making up the receptor binding site are highly conserved but the remainder of the HA molecule is highly mutable. The 16 subtypes of HA recognized currently differ by at least 30% in the amino acid sequence of HA1 and are not cross-reactive serologically. Subtypes may include several variant strains which are only partially cross-reactive in serological assay.

The NA is the second major surface antigen of the virus which, like HA, is an integral membrane glycoprotein. It functions to free virus particles from host cell receptors, to enable progeny virions to escape from the cell in which they arose, and so facilitate virus spread. This activity destroys the HA receptor on the host cell preventing progeny virions reabsorbing to the host cell. The NA is specifically targeted by the NA inhibitor class of antiviral drugs, namely oseltamivir and zanamavir. By inhibiting the NA, these antivirals affect release of virions from infected cells, slowing the spread of the virus and giving time for the host immune response to act. Like HA, the NA is highly mutable, with variant selection driven by host immune pressure. The nine subtypes so far identified in nature are not cross-reactive serologically, although variants within subtypes are partially cross-reactive serologically. The NA inhibitors are effective against all 9 NA subtypes.

There are a number of molecular determinants of influenza virus pathogenicity that include the amino acid residues found at the HA cleavage-activation site, HA receptor specificity, the plasminogen binding ability of NA, specific amino acid changes in PB2 that alter the rate of RNA synthesis, the PB1-F2 ORF and the ability of PB1-F2 to induce apoptosis, and the differing ability of NS1 proteins to counteract the interferon system.

The HA plays a critical role in pathogenicity by mediating virus binding to host cells and fusing the viral and endosomal membranes for viral ribonucleoprotein release into the cytoplasm.

Host specificity is determined by preferential binding of some influenza viruses with sialic acid in the receptor binding site that is linked to galactose by an α2,6- or α2,3-linkage. The receptor binding specificity of human influenza viruses involves α2,6Gal-linkages which is matched by the α2,6Gal-linkages on human epithelial cells in the trachea.

The NS1 protein has been reported to have a number of functions including controlling the temporal synthesis of viral mRNA and viral genomic RNAs, delaying virus-induced apoptosis and avoiding host cell antiviral responses by limiting interferon induction and host T cell activation (Neumann et al. 2009).

In the twentieth and twenty-first centuries there have been four major pandemics in humans, caused by viruses antigenically ‘new’ to the host population (antigenic shift), interspersed with both minor and major epidemics. Pandemic strains generally appear through genetic reassortment. Because vRNA is segmented, genetic reassortment can readily occur in mixed infections with different strains of influenza A viruses. This means that when two viruses infect the same cell, progeny viruses may inherit sets of RNA segments made up of combinations of segments identical to those of either of the parent viruses. This gives a theoretical possible number of 2⁸ (=256) different combinations that can form a complete set of RNA segments from a concurrent infection, although in practice only a few progeny virions possess the correct gene constellation required for viability. The new subtypes of influenza viruses which appeared in humans in 1918 (Spanish influenza), 1957 (Asian influenza), 1968 (Hong Kong influenza), 1977 (Russian influenza) and 2009 (Swine influenza) had several features in common. Their appearance was sudden, they were antigenically distinct from the influenza viruses then circulating in humans, they were confined to H1, H2, and H3 subtypes and, with the exception of the 2009 pandemic virus, the first outbreaks occurred in south east Asia. Phylogenetic evidence suggested that these pandemic strains were derived from avian influenza viruses either after re-assortment or by direct transfer. There is evidence for genetic re-assortment between human and animal influenza A viruses in vivo (Brown et al. 1994; Dawood et al. 2009) and between human influenza viruses (Guo et al. 1992a). The appearance of the H2N2 and H3N2 subtypes was accompanied paralleled by the disappearance from the human population of the previously circulating subtypes, H1N1 and H2N2, respectively. This phenomenon probably occurred in 1918 when emerging H1N1 viruses replaced H3-like viruses. The reasons for the sudden disappearance of previously circulating human strains are unknown at a competition disadvantage but it is possible that the earlier strain is compared with the new strain because it has already elicited widespread immunity in the human population. This may explain the failure of the H1N1 virus to replace H3N2 on its re-emergence in 1977, as a large proportion of the population would have been infected with H1N1 prior to 1957 and retained some immunity. In 2009 H1N1 virus replaced the previously circulating H1N1 whilst the H3N2 continued circulating.

It appears that at frequent but irregular intervals between the major pandemics, variants of pandemic viruses arise which are sufficiently different antigenically to be capable of transmission within the population and thus cause an epidemic. In general, each new ‘epidemic’ variant appears to wane gradually in its ability to find new susceptible hosts and dies out, it is then replaced by the next epidemic variant. These variants arise due to gradual change, i.e. by mutation and selection of the original pandemic virus, and this is termed antigenic drift. Occasionally, the epidemics occurring between pandemics may be sufficiently severe and widespread to mimic a true pandemic. In 1946, an influenza A virus produced worldwide infections that were considered by some to represent a pandemic. However, epidemiological patterns were unlike those of true pandemics, and subsequent antigenic and genetic analyses revealed that the virus was a variant of the H1N1 subtype rather than a representative of antigenic shift.

The phenomenon of antigenic variation by shift and drift in influenza A viruses contrasts with influenza B viruses which show antigenic drift but not antigenic shift, resulting in regular epidemics but not explosive pandemics. Influenza C viruses do not show antigenic drift or shift, apparently only producing sporadic infections.

Two hypotheses have been proposed for the rhythm of occurrence of human influenza A viruses: an influenza circle/cycle or an influenza spiral (Shortridge 1992). The circle/cycle theory suggests there is simply a recycling of H1, H2, and H3 subtypes. The spiral theory presupposed that humans could be infected with all known HA subtypes of influenza A viruses as antibodies to the avian subtypes H4 to H13 were found in serum collected from rural dwellers in the influenza epicentre (Shortridge 1992). It is possible that the hypotheses are not mutually exclusive, as there is no reason that recycling should not occur within the spiral. The pandemic H1N1 virus of 2009 has further complicated these hypotheses since, as a direct descendant of the 1918 H1N1 strains circulated in conjunction with the 1968 H3N2 re-assortant virus as well as the 2009 swine-origin pandemic influenza virus (Shinde et al. 2009; Garten et al. 2009). It seems more likely that with 16 HAs and 9 NAs circulating in an avian reservoir that a number of events need to coincide in order for an influenza virus with pandemic potential to appear. This would involve a unique influenza virus of avian/swine descent to emerge and to reassort with a circulating human-adapted virus. In addition, the herd immunity of different age groups of the human population will vary as exposure to circulating influenza viruses change. This will also drive viral evolution.

Mammalian influenza A viruses replicate primarily in respiratory tract epithelial cells, whereas AI viruses replicate in both the respiratory and the intestinal tracts of birds. Viral shedding in the faeces provides the major mechanism for virus transmission from and between birds. Influenza A viruses capacity to replicate in the lungs is determined by temperature, while replication in the intestinal tract is dependent on pH. Replication of mammalian influenza A viruses is optimal at 33–35°C, reflecting the temperature found in the respiratory tract. AI viruses replicate efficiently at 40–42°C, whereas human influenza viruses do not. In contrast, some swine and equine strains possess the ability to replicate at 42°C, showing intermediate characteristics between avian and human influenza viruses. Furthermore, AI viruses replicate to high titres in the respiratory tract of pigs and can be transmitted readily to other pigs. Similarly, A1 viruses appear to have crossed the species barrier into horses and have been maintained independently of the avian population. The enterotropic avian influenza A viruses are more resistant to low pH, which enables them to pass through the pH barrier in the upper digestive tract of the host. However, a number of influenza A viruses have the potential to replicate in the intestinal tissues of some mammals such as ferrets.

The receptor specificity of the HA differs among influenza viruses and corresponds to the host receptors in the replication site from which the virus was isolated. AI viruses preferentially bind the sialic acid-α-2,3-galactose linkage, while human influenza viruses preferentially bind the sialic acid-α-2,6-galactose linkage on cell surface receptors. Both linkages are found in the epithelial cell site of virus replication lining the pig trachea, in contrast to both birds and humans. The ability of an influenza virus to replicate in avian or mammalian tissues may be genetically linked to the PB2 gene. Studies suggest that the amino acid lysine at residue 627 of the PB2 gene is a determinant of viral pathogenicity in several mammalian species by affecting viral replicative ability. In addition, the amino acid at residue 701 is another virulence determinant, whereas glycosylation of the HA gene appears to control the host range of H1 viruses. It would appear therefore that both viral and host genetic factors determine the tissue tropism of influenza viruses in mammals.

In vitro growth of influenza viruses is usually carried out in 9–11 day old embryonated fowls’ eggs following inoculation of infective material into the allantoic or amniotic cavities. Incubation is carried out according to the virus host’s requirements for 2–4 days, prior to the collection of allantoic/amniotic fluids. In addition, various cell systems have been used for growth, including canine kidney cells, calf kidney cells, human embryonic lung cells, chicken embryo fibroblasts, and conjunctival cells. Organ cultures from fetal and adult trachea have also been used. Whereas embryonated fowls’ eggs are widely used for primary isolation and growth of influenza viruses due to their sensitivity, caution must be exercised in their wholesale application since, with human influenza viruses, genetically distinct variants are selected on passage in the allantoic cavity when compared to those of tissue culture cells, and the egg-adapted variant is not always representative of virus circulating in the human population.

High titres of virus are produced by influenza A infection in bird populations. For example, the huge numbers of ducks congregating on lakes in Canada prior to their migration south and the intestinal site of multiplication of influenza virus in these animals resulted in large doses of virus being excreted into lake water (Webster et al. 1978). Not only has it been shown that infections may be present at such levels to allow virus isolation from untreated samples of lake water (Hinshaw et al. 1979), but also that infectious virus may persist for up to 207 days at 17°C and for even longer periods at 4°C. The infectivity of influenza viruses in water is dependent on the strain of virus, salinity, pH, and temperature of the water. Pigs infected with H1N1 influenza A virus of low virulence may retain live virus in their frozen tissues for up to 3 weeks after slaughter (Romijn et al. 1989). It has been postulated that the reappearance of an H1N1 influenza virus in humans in 1977 (following its disappearance in 1950), was due to reintroduction from a frozen source (Webster et al. 1992).

Influenza viruses are found in a wide variety of mammalian and avian species. Pathogenicity differences among influenza viruses result in the production of a spectrum of clinical diseases that range in severity from fatal systemic disease to mild, sometimes inapparent, respiratory disease. Severity of the disease is also determined by the host species infected and in part by factors such as age, gender, virus dose, environment, and concurrent infections with other pathogens.

In humans, the incubation period varies between 1–5 days, depending on the virus strain and infective dose. Typically, the onset of clinical signs is rapid, characterized by malaise, fever, rhinorrhoea, an unproductive cough, myalgia, and headache. The illness leads rapidly to prostration which usually lasts from 3–5 days, being most severe in children and the elderly. Complications include primary viral or secondary bacterial pneumonia.

The disease in pigs and horses is similar to that in humans. After an incubation period of 1–3 days, disease signs appear suddenly in all or a large number of animals of all ages within a unit. An acute febrile, respiratory disease is characterized by fever, apathy, anorexia, coughing, sneezing, nasal discharge, conjunctivitis, a low mortality rate, and a rapid recovery. Secondary bacterial infections in both pigs and horses can often increase the severity of the illness and may result in complications such as pneumonia.

In birds the disease signs can vary considerably. Typical clinical signs of highly pathogenic AI in chickens or turkeys include decreased egg production, respiratory signs, rales, excessive lacrimation, sinusitis, cyanosis of unfeathered skin especially combs and wattles, oedema of head and face, ruffled feathers, diarrhoea, nervous disorders, and high mortality. Nonpathogenic AI viruses may replicate in the epithelial cells of the respiratory tract and the intestine of birds without inducing signs of disease, but virus may be shed at high concentrations in the faeces. Exacerbative conditions, including infection with other organisms, may result in AI viruses which are normally not pathogenic causing severe disease in infected birds.

Generally, influenza virus infections of other animals such as ruminants and sea mammals result in subclinical disease. However, H7N7, H4N5, and H3N3 influenza A viruses were associated with a high mortality in harbour seals at Cape Cod, USA in 1979 (Lang et al. 1981), 1982 (Stuart-Harris et al. 1985) and 1991 (Callan et al. 1995), virus being isolated from lung and brain of dead animals. Mustelids may be more susceptible to influenza infections; in addition to the historical laboratory infections of ferrets, outbreaks of severe respiratory disease with 100% morbidity and 3% mortality in commercial mink in Sweden were considered due to infections with influenza virus of H10N4 subtype. Isolates made from the mink showed close genomic homology with H10N4 viruses circulating concomitantly in avian species (Berg et al. 1990). In all these cases the epidemics tended to be self-limiting, and the newly introduced viruses did not appear to be maintained in sea mammals or mink.

The respiratory pathology following infection with influenza virus is difficult to define precisely, since it is frequently complicated by the effects of secondary bacterial infection. Severe damage to the epithelium of the respiratory tract is the main feature in infections of mammals with influenza virus. The resulting damage to the epithelium facilitates secondary infection by bacterial respiratory pathogens.

In humans there is typically a rhinitis followed by tracheobronchitis and, infrequently, an interstitial pneumonitis. The disease process usually damages the respiratory tract from the nose to the small bronchi, but rarely damages alveolar cells. However, in some cases influenza A virus infection results in gross lung lesions which are patchy and randomly distributed throughout the lobes. The altered lung areas are depressed and consolidated, dark red or purple red in colour, contrasting sharply with normal lung tissue (Morens et al. 2010).

In typical infections of pigs the bronchi and bronchioli are dilated and filled with exudate. Bronchial and mediastinal lymph nodes are usually hyperaemic and enlarged. Histologically, there is widespread degeneration and necrosis of the epithelium in the bronchi and bronchioli. The lumen of bronchi, bronchioli, and alveoli are filled with exudate containing desquamated cells and neutrophils progressing to mainly monocytes. Furthermore, dilatation of the capillaries and infiltration of the alveolar septae with lymphocytes, histiocytes, and plasma cells occurs. Widespread interstitial pneumonia and emphysema accompany these lesions, although the severity of the former is dependent on the infecting strain (Brown et al. 1993a).

Significant pathological changes in other organs have not been consistently observed among infected mammals. Infections of birds with highly virulent avian viruses are characterized by haemorrhagic, necrotic, congestive, and transudative changes. Haemorrhagic changes are frequently severe in the oviducts and intestines. Encephalitis may develop in the cerebrum and cerebellum, especially in broilers. Alterations to myocardial tissues have been observed following infection with highly pathogenic strains.

Laboratory methods for detecting influenza virus and other respiratory virus infections have been revolutionized since the start of the twenty-first century. Classical techniques are still used in certain settings, such as haemagglutination inhibition, microneutralization and virus isolation. However, molecular based rapid diagnostic methods are being used around the world including reverse transcriptase and real time polymerase chain reaction (PCR) assays as well as microarrays to detect influenza virus RNA as well as type the virus; automated sequence analysis to determine whether that virus contains any antiviral resistance mutations as well as molecular epidemiological analysis using high throughput automated sequencing (Wang and Taubenberger 2010).

Clinical diagnosis of infection with influenza virus is only presumptive. The rapid confirmation of influenza requires that a respiratory sample should be tested for the presence of influenza virus RNA. Alternative classical methods that take longer periods of time to make a diagnosis include virus isolation in cell culture or testing a blood sample for influenza virus antibody, essentially demonstrating a recent infection in paired samples or a high titre antibody in a recently collected sample. Molecular diagnostic assays have revolutionized both the diagnosis and surveillance of respiratory virus infections in terms of sensitivity of detecting various infectious agents, especially influenza. Although the clinical picture includes a high temperature, myalgia and a variety of respiratory symptoms and signs, the pandemic influenza A H1N1 virus infections in 2009 demonstrated that a number of other respiratory viruses can also cause similar symptoms even when the case definition is carefully delineated. In addition to acute disease there may be subclinical infection or atypical courses of infection such as in a partially immune population.

Generally, the best material for viral RNA detection using molecular based methods such as real time PCR assays or virus isolation is nasal mucus from mammals and faecal samples from birds. These are collected using sterile swabs, which are immediately suspended in transport medium, i.e. 40% glycerol, 60% saline, to prevent them drying out. Some laboratories have used specimen tubes that contain viral lysis buffer that ensure the stability of the sample, although these samples are unhelpful if virus isolation is required. Samples should be collected as soon as possible during the acute phase of the disease. In addition, tissues from the respiratory tract of mammals and from respiratory, intestinal, and systemic organs of birds are suitable for viral RNA detection or virus isolation. Tissues may be homogenized in saline containing antibiotics and antimycotics and clarified to obtain a clear supernatant. The most suitable, easily available and reliable host system for the isolation of influenza viruses is 9–11 day old embryonated fowls’ eggs.

Various cell cultures may also be used for the isolation of influenza viruses, but the Madin Darby canine kidney cell line is most frequently used for influenza isolation from humans and other mammals. Usually it is necessary to add trypsin to the growth medium, as a conditioning factor for the cleavage of the HA and the production of infectious virus.

After incubation at 35°C (mammalian influenza) or 37°C (avian influenza) for 2–4 days, the allantoic/amniotic or cell culture fluids are collected and tested for the presence of HA in the haemagglutination test using chicken red blood cells. Positive haemagglutination is presumptive for the presence of an influenza virus for mammals, but avian species are commonly infected with Newcastle disease and other paramyxoviruses. Initial identification of a virus is performed by the immunodiffusion test with specific antisera to the nucleoprotein or matrix protein of the three types of influenza virus. This confirms the isolate as an influenza virus of type A, B, or C, and distinguishes it from all other agents that exhibit haemagglutination. Further characterization of influenza A viruses is carried out to identify the antigenic nature of the surface antigens, HA and NA, in haemagglutination inhibition (HI) and neuraminidase inhibition (NI) tests, respectively, using a panel of monospecific antisera for each of the 16 HA and 9 NA types. The specific inhibition of HA and NA permits subtype identification of the influenza A virus.

Classical techniques include the use of serology for diagnosis which can be particularly useful when virus shedding is brief and is of low titre, as is often the case with respiratory viruses or when an animal or individual is no longer in the acute phase of the disease. The HI test (Palmer et al. 1975), may be used to diagnose influenza and offers the advantages of being relatively simple to perform, sensitive, easily adaptable, and inexpensive. The single radial immunodiffusion test is an alternative assay which is also inexpensive and simple to perform. Paired sera taken in the acute stage of illness and approximately 2–3 weeks later during convalescence are required for diagnosis, in order to demonstrate an increase in specific antibodies. In epidemic situations when influenza is suspected and paired sera are unavailable, a rapid diagnosis can often be made by examining single serum specimens from selected individuals for elevated levels of influenza antibody.

The serum of many species, particularly mammals, contains inhibitory substances that may interfere with the specificity of HI and other tests. Various treatments of sera have been suggested to remove these inhibitors, possibly the best and most widely used being incubation with receptor destroying enzyme. In addition to non-specific inhibitors of the HA, some sera contain non-viral substances that may agglutinate certain species of red blood cells used in the HI test. These substances may be removed by pre-treating the serum with erythrocytes to be used in the HI test.

Overall, molecular methods are the mainstay for the analysis of influenza virus infections in terms of rapid diagnosis, typing, determining antiviral susceptibility and for epidemiological purposes.

The two classes of antiviral drugs used in both treatment and prophylaxis target the M2 and NA envelope proteins.

The adamantanes, amantadine and rimantadine, are M2 ion channel blockers and are effective against influenza A viruses only. These drugs date back to the early 1960s and 1990s respectively. They block the efflux of hydrogen ions due to the change in pH as they are basic compounds, and interfere with intracellular virus uncoating. They reduce the duration of fever by 24 hours and prevent 60–70% of influenza A infections if used as prophylaxis. They are associated with neuropsychiatric side effects, especially in the elderly if used in higher doses.

Amantidine-resistant influenza A virus is widespread, having first been detected in 1981. In the USA, the incidence of adamantane resistance rose to 92% in 2005. This is due to M2 gene mutations and nearly 80% of resistant viruses have a serine to asparagine mutation at codon 31. Naturally occurring influenza A viruses are essentially quasispecies of susceptible and resistant strains. The latter dominate within 3 days of starting amantadine treatment and have similar virulence to the wild type virus. The M gene of pandemic influenza H1N1 virus is similar to the M gene in the Eurasian swine virus, which confers resistance to both amantadine and rimantadine (Jefferson et al. 2006; Suzuki et al. 2003).

The NA inhibitors, oseltamivir and zanamivir, were made available in 1999 and are effective against both influenza A and B viruses. Oseltamivir (Tamiflu) is more widely used as it can be taken orally as opposed to zanamivir (Relenza) which is administered by inhalation. NA inhibitors interfere with the release of new influenza virions from infected cells, preventing the infection of new cells. Both drugs are effective in reducing the median time to alleviating influenza symptoms by up to 24 hours. The benefit of prophylaxis and treatment has been shown when these drugs are given within 48 hours of symptom onset in previously well adults. However, further analyses are awaited to demonstrate effectiveness in other clinical settings. With respect to post-exposure prophylaxis, an 80–90% reduction in influenza incidence has been reported.

Zanamivir is associated with cough, bronchospasm and death in individuals with underlying respiratory disease and is contraindicated. Oseltamivir is associated with gastrointestinal disturbances in 10% of individuals, especially nausea and vomiting.

Oseltamivir resistance has been widely reported and usually involves the histidine to tyrosine at codon 274 in N2 nomenclature or H275Y in N1 nomenclature mutation in the NA gene. In 2008, up to 99% of seasonal H1N1 isolates in the USA were oseltamivir resistant. Transmission of oseltamivir resistance has occurred in the absence of direct selective drug pressure without affecting either virulence or replicative capacity.

In general, zanamivir retains full inhibitory activity against several NA subtypes in the presence of oseltamivir resistance mutations. However, although zanamivir resistance is rare, it has been demonstrated after treating immunocompromised individuals. Most of the 2009 pandemic H1N1 isolates remained susceptible to the NA inhibitors. However, there is the potential, with the seasonal and pandemic influenza viruses co-circulating, for incorporation of the H274Y mutation leading to oseltamivir-resistant pandemic H1N1 virus infections (Hurt et al. 2006, 2009; Moscona 2005; Cooper et al. 2003; Jefferson et al. 2006, 2009; Weinstock and Zuccotti 2009; Moscona 2005, 2009; Dharan et al. 2009; Lackenby et al.  2008 ).

Oseltamivir and zanamivir have been made available as intravenous preparations (Gauer et al. 2010). Peramivir is another NA inhibitor that has been developed and can be given intravenously (Barroso et al. 2005). Finally, laninamivir is a long-acting NA inhibitor that is structurally related to zanamivir, but has a longer half-life.

Novel molecular targets involved in different steps in the influenza virus life cycle have been investigated (Hayden 2009). Drugs that have been studied include favipiravir that is converted into a nucleotide analogue that inhibits influenza virus RNA polymerase (Smee et al. 2009; Furuta et al. 2009). In addition, fludase is a sialidase catalytic domain/amphiregulin glycosaminoglycan binding sequence fusion protein. The mechanism of action is unique as it selectively cleaves sialic acid receptors from the host cells, so the influenza virions cannot bind (Moss et al. 2010).

Finally, another form of treatment that has a historical basis involves the use of hyperimmune plasma. This was used in the 1918 Spanish influenza pandemic and was made from blood collected from convalescent human volunteers and given to patients with severe influenza infections. Successful outcomes were reported. There were reports that in individuals with severe pandemic H1N1 infections, the IgG2 subclass deficiency could be corrected by hyperimmune plasma infusions that had been collected from individuals with pandemic H1N1 infection or from vaccinated donors (Gordon et al. 2010). This is a potential therapeutic adjunct to conventional antiviral treatment in the treatment of severe cases of pandemic H1N1 infection (Moss et al. 2010).

Antibiotics and other antibacterial agents do not affect the viral infection, but may sometimes be used to prevent complications such as bacterial co-infection. Specific control measures include the use of antiviral drugs and vaccination. General prophylactic measures in mammals and birds are based mainly on preventing the introduction of influenza viruses of wild aquatic birds into domestic pig herds and poultry flocks. Infected herds or flocks are kept warm and free from stress for a more rapid recovery.

In humans the effects of infection with influenza viruses in a population are most easily measured by comparing excess overall mortality with ‘pneumonia-influenza’ death rates, and are used to indicate the extent of an epidemic rather than the lethality of the virus. It is estimated that the 1918 pandemic resulted in 50–100 million deaths worldwide in all age groups. Subsequent pandemics in 1957 and 1968 showed dramatically increased excess mortality, but the effects compared to the 1918 pandemic were much reduced, possibly reflecting the availability of antibiotics in preventing deaths as a result of secondary bacterial infections. The combined pandemics of 1957 and 1968, in the USA, accounted for approximately 98, 000 excess deaths; however, the epidemics from 1957 to 1975, excluding the pandemic years, accounted for over twice that number of excess deaths (Dowdle 1976), indicating that epidemic influenza occurring as a result of antigenic drift is a significant ‘killer’ disease in humans.

In most age groups, influenza infections produce high morbidity, but recovery is usually rapid and uneventful. Mortality is highest in the elderly and the very young, frequently accounting for 90% of all mortality associated with influenza virus infections. If a vaccine provided 100% protection, about 80% of influenza-related deaths could be prevented by vaccinating all people above 70 years of age (Sprenger et al. 1993).

In other mammals the disease usually produces a short illness, characterized by low mortality, high morbidity, and rapid recovery, but varies with the infecting strain of virus and the affected species. An H7N7 influenza A virus was associated with a high mortality in harbour seals in the USA in 1980, but was apathogenic for chickens and turkeys. A novel strain of equine influenza virus (H3N8) which emerged in horses in China in 1989 was associated with a high morbidity and relatively high mortality of up to 20%. The virus appeared to originate from birds (Guo et al. 1992b). The introduction of an avian-like H1N1 virus into an immunologically naive pig population resulted in a large number of disease outbreaks characterized by high morbidity but low mortality (Brown et al. 1993b).

Most influenza viruses infecting birds produce asymptomatic disease. Outbreaks in poultry due to highly pathogenic AI viruses are rare, but when the disease does occur it may result in up to 100% mortality, often with few clinical signs preceding sudden death.

Phylogenetic studies of influenza A viruses have revealed species-specific lineages of viral genes and have demonstrated that the frequency of interspecies transmission depends on the animal species. In the early 1970s, the WHO initiated long-term global studies on the influenza viruses of mammals and birds to determine the diversity of influenza A viruses in nature and whether it was possible to isolate a future pandemic strain from them in advance of its appearance in humans.

To the present day, a large number of viruses have been isolated from a wide variety of birds and a range of terrestrial and sea mammals. These can all be grouped into 16 HA subtypes and 9 NA subtypes, suggesting there may be a limited range of antigenic subtypes in nature. However, it has been reported that phylogenetic analyses of the 1918, 1957 and 1968 pandemic viruses suggested that all evolved undetected in an intermediate mammalian host well before human infections were seen (Smith et al. 2009). In addition, systematic virological surveillance of swine influenza virus infections in a Hong Kong SAR abattoir, over a decade revealed a high degree of reassortment (Vijaykrishna et al. 2010).

Intraspecies transmission of influenza A viruses does occur, but is infrequent, and occurs most readily between host species that are closely related in which transmission is sustained. Following transmission, the influenza A virus must adapt to the new species before efficient and high level replication occurs. There is an association between efficient replication and expression of virulence.

Although a wide range of animal species are susceptible to influenza A virus infections, three groups of animals appear to be more important in terms of numbers and the epidemic/endemic nature of the disease than other animals: these are birds, pigs, and horses.

Influenza viruses infect many avian species naturally a great variety of birds, including wild birds, captive caged birds, domestic ducks, chickens, turkeys, and other domestic poultry (Perdue and Swayne 2004).

Viruses have been isolated from species of wild bird covering all the major families of birds. This has led to the findings that nonpathogenic AI viruses are ubiquitous, particularly in aquatic birds, and that all of the different subtypes of influenza A viruses (H1 to H16 and N1 to N9) are perpetuated in aquatic birds, particularly migrating waterfowl. Furthermore, phylogenetic studies have revealed that aquatic birds are probably the source of all influenza viruses in other species.

The frequency at which viruses have been isolated from samples taken from waterfowl has varied considerably. Hinshaw et al. (1980) contrasted the frequent isolation of virus from ducks congregated on lakes in Alberta, Canada with the much lower isolation rates obtained from birds on migration. Some of the factors that governed whether or not waterfowl were likely to be infected and excrete virus were the age of the bird, the geographical location relative to migration, the time of year, the species, and the characteristics of a particular virus. Each year, waterfowl congregate in extremely large flocks, usually on lakes, before migratory flights are undertaken. At this stage, viruses may spread easily to susceptible birds on the crowded lakes. Isolation rates from juvenile ducks may exceed 60%. The importance of waterfowl is not only in the antigenic diversity and size of virus pools they harbour, but also the rapid dissemination of these viruses around the world due to the migratory nature of these birds.

In wild ducks, influenza viruses replicate mainly in the intestinal tract and are excreted in high concentrations in the faeces. Many birds, particularly juveniles, are infected by the virus shed into the lake water; however, viral genetic information does not persist in the individual after clearance of infectious virus, which is usually 5–7 days after infection. Certain subtypes of influenza virus predominate in wild ducks along a particular flyway, but the predominant virus differs from one flyway to another from year to year. Studies of ducks and swans from Siberia wintering in Japan have shown an influenza isolation rate during the winter months which varied from 0.5–9%, year to year.

Influenza viruses of a variety of HA and NA subtypes have also been isolated from wild waterfowl in other parts of the world, including Russia, southern China, western Europe, and Australia demonstrating the worldwide distribution of avian influenza virus gene pools in nature. Phylogenetic studies have indicated that influenza viruses from Eurasia and Australia are genetically distinct from those in North America. These studies indicated that the incidence and prevalence of influenza subtypes will vary due to physical barriers which prevent intermixing of their hosts. Studies by Sharp et al. (1993), suggested that whereas wild ducks perpetuated some influenza A viruses, they did not act as a reservoir for all such viruses. It has been suggested that the remainder of the influenza gene pool is maintained in shorebirds and gulls, from which the predominant isolated influenza viruses are of a different subtype to those isolated from ducks. Circumstantial evidence suggested that initial outbreaks in domestic poultry most often occurred as the result of spread from wild birds, although there were several reports of influenza viruses transmitted from pigs to turkeys. As a consequence, considerable antigenic variation has been seen in disease outbreaks in domestic poultry.

AI viruses are classified into low (LP) and high pathogenicity (HP) viruses based on virulence in chickens using an intravenous pathogenicity test. The LP viruses mostly cause respiratory or reproductive system infections with low mortality rates and do not meet the definition of an HPAI virus. The latter may produce more than 75% mortality and include any H5 and H7 AI viruses that have a haemagglutinin proteolytic cleavage site compatible with a HPAI virus. AI viruses are maintained as LPAI viruses in the wild bird reservoir. Following transfer and circulation in domestic poultry, some H5 and H7 LPAI viruses have mutated to HPAI viruses.

Influenza viruses have been isolated less frequently from feral passerine birds than from waterfowl. Studies following a highly pathogenic H7N7 outbreak in chickens in Australia concluded that there had not been significant spread to feral birds, although virulent virus was isolated from a starling found on the affected farm (Morgan and Kelly 1990). Captive, caged, and pet birds may also have a role to play in the propagation and dissemination of influenza viruses. Monitoring of such birds throughout the world has resulted in the isolation of many viruses.

It may be concluded that enormous pools of both genetically and antigenically diverse influenza viruses exist within the bird population, and provide the source of virus for the mammalian population.

In early 1997, an H5N1 influenza virus known to infect only birds caused an outbreak involving 18 people in Hong Kong, 6 of whom died. A high mortality rate had been seen in chickens infected with H5N1 influenza in three farms in Hong Kong. It was subsequently shown that individuals in close contact with the index case or with exposure to poultry in the live market were at risk of being infected (Yuen et al. 1998; Claas et al. 1998; Subbarao et al. 1998).

Control measures to reduce exposure included culling all poultry in Hong Kong, segregating water fowl and chickens and introducing import control measures for chickens. Although the outbreak occurred, successful controlled H5N1 virus outbreaks occurred in 9 Asian countries in 2003, causing large-scale outbreaks in poultry and at least 7 fatal cases of human infection in three countries (Sims et al. 2003).

Furthermore, between February and May 2003, a fowl plague outbreak due to HPAI virus of subtype H7N7 occurred in the Netherlands. This was closely related to LPAI virus isolates from wild ducks but was isolated from chickens. The same virus was detected subsequently in 89 people, most of whom developed conjunctivitis having handled affected poultry. Three of the 89 were family members whom had no direct contact demonstrating human to human transmission. Influenza-like illnesses were generally mild, but there was a fatal case of pneumonia and acute respiratory distress syndrome (Koopmans et al. 2004; Fouchier et al. 2004; Elbers et al. 2004).

Between November 2003 and June 2008, HPAI H5N1 viruses caused fatal human infections in 296 of 500 confirmed cases in 15 countries in Asia, Africa, the Middle East, and Europe (WHO (2009) http://www.who.int/csr/disease/avian_influenza/country/cases_table_2009_04_17/en/index.html), (Nguyen-Van-Tam et al. 2005; Peiris et al. 2004).

The HA genes of these viruses isolated from humans were consistent with those viruses that infected domestic poultry in southern China. LPAI H5N2, H4N6 and H9N3 viruses were also detected in apparently healthy geese and ducks (Webster et al. 2002).

HPAI H5N1 viruses isolated from domestic waterfowl in 2001 and 2003 were not the progenitors of the viruses that caused widespread outbreaks in poultry and human infections in Vietnam in 2003–2004. The latter H5N1 viruses were independently introduced into Vietnam (Jadhao et al. 2009).

H1N1 and H3N2 influenza A viruses have been widely reported in pigs, frequently associated with clinical disease. These include classical swine H1N1, avian-like H1N1, and human and avian-like H3N2 viruses (Table 30.3). Swine influenza has remained in the pig population and has been responsible for one of the most prevalent respiratory diseases in pigs.

Swine influenza is related to the movement of animals from infected to susceptible herds, clinical disease generally appears with the introduction of new pigs into a herd. Once a herd is infected, the virus is likely to persist through the production of young susceptible pigs and the introduction of new stock. Outbreaks of disease occur throughout the year but usually peak in the colder months. Infection with classical swine H1N1 influenza virus is frequently subclinical, and typical symptoms are seen often in only 25–30% of a herd. Blaskovic et al. (1970) showed that classical swine H1N1 influenza virus was excreted from one infected pig for over 4 months, although 7–10 days is more usual. Continuous circulation of swine influenza viruses within a herd without the apparent need for an intermediate host has been shown by the isolation of virus from a herd all the year round.

Influenza A viruses causing clinical disease reappeared in European pigs in 1976, with the introduction of classical swine H1N1 influenza virus to Italy from North America. Although usually regarded as an endemic disease, influenza infections of pigs may result in epidemics when introduction of virus to an immunologically naive population occurs. Great Britain had remained free of avian-like H1N1 virus until 1992 when respiratory disease was seen spreading rapidly throughout the country as a result of infections with an H1N1 virus related to, but antigenically distinguishable from, the prototype strains of avian-like H1N1 viruses (Brown et al. 1993b).

Human H3N2 influenza A viruses related to a human strain from 1973, circulated in European pig populations long after their disappearance from the human population.

There is good evidence that genetic reassortment can occur in nature between influenza A viruses in pigs, but this has not resulted in new epidemics in the pig population in which it has occurred. Influenza A H1N2 viruses, derived from swine H1N1 and H3N2 viruses, have been isolated in Japan (Sugimura et al. 1980) and France (Gourreau et al. 1994). Phylogenetic analyses of human H3N2 viruses circulating in Italian pigs revealed that genetic re-assortment had been occurring between avian and human-like viruses since 1983. The unique co-circulation of influenza A viruses within European swine may lead to pigs serving as a mixing vessel for reassortment between influenza viruses from mammalian and avian hosts, with unknown implications for both humans and pigs. Further evidence for influenza virus reassortment in the pig is provided by the isolation of an H1N7 virus from pigs in England, apparently derived from human and equine viruses, and the isolation of an H1N2 virus from pigs in Great Britain, apparently derived from human and swine viruses (Brown et al. 1995a). Unlike H1N2 viruses detected elsewhere, this H1N2 appeared to spread widely within pigs in Great Britain.

Up to 1998 in the USA, swine classical H1N1 strains were the dominant viruses circulating in pigs. Since that time, new swine H3N2 strains have been detected and resulted from a double re-assortment of swine classical H1N1 with the PB1, HA, and NA segments from a human H3N2 strain, and a triple reassortment of swine classical H1N1 with the PB1, HA, and NA segments of a human H3N2 strain and the PB2 and PA segments of avian lineage. H1N2, H3N1, H2N3, H4N6, H5N1 and other subtypes have been isolated in pigs around the world as a result of inter-host re-assortments of human and/or avian viruses.

Reassortant H2N3 viruses were characterized having been isolated from pigs with respiratory disease from two farms in the USA, a subtype not previously reported in swine (Ma et al. 2007). These H2N3 reassortant viruses contained genes derived from avian and swine influenza viruses. The virus was able to replicate in pigs, mice, and ferrets and was transmitted among pigs and ferrets.

This was an important finding as they belonged to the H2 subtype as did the 1957 human pandemic strain that disappeared in 1968. Therefore, from a public health perspective, a new generation of people would have little pre-existing immunity to this subtype. Furthermore, they were circulating in swine that could select for mammalian influenza viruses, they had receptor binding site changes associated with increased affinity for α2,6Gal-linked sialic acid viral receptors and could replicate and transmit in swine and ferrets.

The original source of the H2N3 virus was unclear, but it was thought that both farms used surface water collected in ponds for cleaning the barns and watering animals. Therefore, the avian virus may have infected the pigs by direct contact with the contaminated surface water.

Reassortment has also been demonstrated between avian- and human-like influenza viruses in Italian pigs.

Pigs serve as major reservoirs of H1N1 and H3N2 influenza viruses and are often involved in interspecies transmission of influenza viruses. The maintenance of these viruses in pigs and the frequent introduction of new viruses from other species could be important in the generation of pandemic strains of human influenza.

Equine influenza virus (EIV) infection, first isolated in 1963, causes a common disease of horses throughout the world. Apart from rare reports of isolations or serological evidence of infection with other subtypes, only two subtypes of influenza A virus, H7N7 and H3N8, have been identified as infecting and causing disease in horses. It is possible that H7N7 viruses have disappeared largely from the horse population as H3N8 is apparently the only subtype currently circulating in the horse population. Equine influenza outbreaks due to infection with H3N8 virus have been seen in South Africa, India, China, Hong Kong, and Nigeria. Outbreaks in China have been due to both conventional strains of H3N8 virus, and viruses which, although they contained the same surface antigens as the other equine viruses of this subtype, had genetic features that were avian-like, indicating that the virus had been introduced from birds. In North America and Europe, two distinct groups of H3N8 (equine-2) virus have been detected. Sporadic outbreaks typically occur and may in part be due to antigenic drift, which may compromise the efficacy of the available vaccines. EIV has also jumped the species barrier and become established as a respiratory pathogen of dogs (Murcia et al. 2010).

In May 1993, a severe epidemic of respiratory disease spread throughout horses in China due to an H3N8 influenza A virus whose gene segments were derived entirely from classical equine-2 influenza viruses, closely related to an H3N8 equine influenza virus isolated in Sweden in 1991. These results demonstrated that European equine H3N8 influenza viruses had been transmitted to China. Swine influenza virus surveillance during 2004–2006, showed that two strains of H3N8 influenza viruses isolated from pigs in central China were of equine origin and were closely related to European equine H3N8 influenza viruses dating back to the early 1990s (Tu et al. 2009).

Influenza A viruses of HA subtypes H1, H3, H4, H7, and H13 have been isolated from dead and dying seals and whales (Hinshaw et al. 1986). These sea mammals were probably infected from the faeces of birds, shed into the water at communal gathering sites (Shortridge 1992), as it is known that these viruses were of avian origin (Callan et al. 1995). Influenza viruses have been isolated from mink raised on farms. These H10N4 viruses were of avian origin (Berg et al. 1990), caused systemic infection and disease in the mink and spread to contacts. In all these cases the epidemics tended to be self-limiting, and the newly introduced viruses did not appear to be maintained in sea mammals or mink.

Between May and December 2007, HPAI-origin canine influenza A viruses (H3N2) spread across South Korea. The viruses shared more than 97% nucleotide sequence homology, suggesting that whole viruses were transmitted directly from birds to dogs. These viruses were also transmitted experimentally to beagles by direct contact, showing that interspecies transmission of AI viruses was possible (Song et al. 2009).

In January 2004, an outbreak of respiratory disease occurred in 22 racing greyhounds at a Florida racetrack in the USA. Dogs suffered a mild illness with fever and cough or died with a suppurative bronchopneumonia with pulmonary haemorrhage, a 36% case-fatality rate. From June to August 2004, respiratory disease outbreaks occurred at 14 tracks in 6 states with a combined population of ∼10,000 racing greyhounds. Virological, serological, and molecular evidence for interspecies transmission of an entire equine influenza A (H3N8) virus was reported (Crawford et al. 2005). Unique amino acid substitutions in the canine virus HA, coupled with serological confirmation of infection of dogs in multiple states in the USA, were correlated with sustained circulation of the virus in the canine population.

Influenza epidemics in humans can occur following infection with influenza A or B viruses, but pandemics result only from infection with influenza A virus. Influenza A viruses of H1, H2, and H3 subtypes have been associated with infection in humans since the mid-nineteenth century. Pandemics generally occur following antigenic shift, whereas epidemics result from antigenic drift. The 2009 pandemic H1N1 virus was an exception although it does depend on the definition of the term pandemic. Does ‘pandemic’ refer to a global outbreak alone or also infer high morbidity and mortality?

Antigenic shifts have resulted in new subtypes of human influenza A viruses appearing in 1918 (H1N1); in 1957, when the H2N2 subtype replaced the H1N1 subtype; in 1968, when the H3N2 virus appeared, replacing the H2N2 subtype; and in 1977 when the H1N1 virus reappeared. In the last case the reappearance of H1N1 did not result in the replacement of H3N2 viruses and both continue to circulate. Pandemic strains appear to arise at one focus and spread rapidly worldwide. Each of the new subtypes since 1957 first appeared in China and spread across all continents as a result of the available, fully susceptible world population. For example, the H1N1 virus that reappeared in May 1977 in humans in northern China, had spread to much of south east Asia by November, and by February 1978 it was already present in Europe and the USA. Within 9 months, this virus had been distributed almost worldwide (Kendal 1987).

Every few years a new antigenic variant of the prevailing subtype appears that is capable of spreading in the population and causing a significant epidemic. In temperate climates, epidemics nearly always start at the beginning of winter. They are of varying severity dependent on the infecting strain but are characterized by rapid spread, often infecting between 30–40% of the population of the affected area. Epidemic strains usually arise at one focus and spread rapidly worldwide. After their rapid start, the epidemics tend to end no less abruptly, often within several weeks at the local level, or within 3 months nationwide. The sudden cessation of epidemics, often when there are still many susceptible individuals in the population, has not been explained.

Epidemics occur worldwide in the human population due to infection with influenza A viruses of H1N1 or H3N2 subtype or with influenza B virus. Rarely, infections can occur with re-assorted influenza A viruses such as H1N2 and H3N1 although spread of such viruses appears to have been very limited.

Occurrence rates for influenza virus infection are highest in children. This is partly attributable to the increased immunity in older individuals that results from prior infections with related viruses. Children and young adults play an important role in the dissemination of virus into the community. When adults, especially those 65 years of age or older, are exposed to influenza virus in a setting such as a nursing home, the infection rate can be as high as in younger persons, but with the potential for more severe consequences. Death rates from influenza during an epidemic are invariably high enough to affect the overall death rate and it is chiefly in the elderly that such deaths occur. Other groups at risk include people with chronic disorders of the respiratory or cardiovascular systems, children with asthma, immunosuppressed patients or those with metabolic disorders, persons infected with human immunodeficiency virus, pregnant women, regular foreign travellers and those involved in the health care of some of the aforementioned groups.

The detection of vast pools of influenza viruses of many different subtypes among animals, particularly aquatic birds gave considerable impetus to research aimed at determining where new subtypes, particularly those that cause pandemics, emerge. Many theories have been suggested, of which the most widely accepted is that by adaptation, involving genetic reassortment, virus transmission from other animals to humans occurs which results in an antigenically novel virus with the ability to infect and spread in humans. Following genetic reassortment, viruses may arise that possess the necessary genes to enable infection of humans but may have surface antigens new to the host immune system.

Genetic and biochemical studies have shown that the 1957 and 1968 pandemic viruses arose by genetic reassortment. The HA, NA and PB1 genes of the 1957 Asian H2N2 strain were from an avian virus and the remaining five genes from the preceding human H1N1 strain (Kawaoka et al. 1989). The 1968 Hong Kong H3N2 strain contained HA and PB1 genes from an avian donor and the NA and other five genes from the Asian H2N2 strain. A diagrammatic representation of the theoretical origin of influenza A viruses in humans is shown in Fig. 30.1.

The pig has been the leading contender for the role of intermediate host for reassortment of influenza A viruses. Pigs are the only mammalian species which are domesticated, reared in abundance and are susceptible to, and allow productive replication of, avian and human influenza viruses. This susceptibility is due to the presence of both α2,3- and α2,6-galactose sialic acid linkages in cells lining the pig trachea, which can result in modification of the receptor binding specificities of avian influenza viruses from α2,3 to α2,6 linkage; thereby providing a potential link from birds to humans. Furthermore, it has been shown that humans occasionally contract influenza viruses from pigs. The internal protein genes of human influenza viruses share a common ancestor with the genes of most swine influenza viruses. Also, the pig has a broader host range concerning the compatibility of the NP gene of viruses derived from other species (Scholtissek et al. 1985).

AI viruses which do not replicate in pigs can contribute genes that generate re-assortants when co-infecting pigs with a swine influenza virus. Evidence for the pig as a mixing vessel of influenza viruses of non-swine origin has been demonstrated in Europe by Castrucci et al.(1993), who detected a reassortment of human and avian viruses in Italian pigs. In addition, human and equine influenza viruses resulting in re-assortant viruses (Brown et al. 1994) or from human and swine viruses (Brown et al. 1995a) have been isolated from pigs in Great Britain.

Alternatively, new pandemic viruses could occur in the human population if an avian strain or a strain from another mammal became infectious for humans. Phylogenetic evidence supports this mechanism for the appearance of the Spanish influenza virus (H1N1) in 1918. Analyses of the NP gene, which is associated with host specificity, of human, swine, and avian H1N1 viruses reveals that the classical swine viruses and the contemporary human viruses probably evolved from a common avian ancestor prior to the appearance of the 1918 human pandemic strain (Gorman et al. 1991). Furthermore, avian-like H1N1 viruses circulating in European pigs since 1979 were implicated as the precursors of the next human pandemic virus (Ludwig et al. 1995). In addition, avian H1N1 viruses antigenically and genetically related to, but distinct from European avian-like swine viruses have been detected in pigs in south east Asia since 1993 (Webster et al. 1996).

Finally, the appearance of pandemic virus may be in fact be the re-emergence of a virus which may have caused an epidemic many years earlier. The appearance of Russian influenza (H1N1) provided support for this concept. The virus that reappeared in China in 1977 and spread subsequently to all parts of the world was identical in all of its genes to the virus which caused a human influenza epidemic in 1950 (Nakajima et al. 1978). Webster et al. (1992) suggested that this virus was most likely reintroduced to humans from a frozen source and Shoham (1993) proposed a biotic mechanism for the preservation of influenza viruses. Influenza viruses of the H3N2 subtype persist in pigs many years after their antigenic counterparts have disappeared from humans and therefore present a reservoir of virus which may in the future infect a susceptible human population. Pandemic strains may also be antigenically conserved in the avian reservoir, since counterparts of the Asian pandemic strain of 1957 circulated with increasing prevalence in wild ducks, domestic fowl, and live bird markets, coming into closer proximity to susceptible human populations.

The majority of pandemic strains have originated in China, raising the possibility that this region is an influenza epicentre (Shortridge and Stuart-Harris 1982; Shortridge 1992). In the tropical and subtropical regions of China, influenza occurs all year round. In China, influenza viruses of all subtypes are prevalent in ducks and in water frequented by ducks. The agricultural practices provide that there is close contact between domestic ducks, pigs, and humans, thereby presenting the opportunity for interspecies transmission and genetic exchange among influenza viruses, with the pig acting as an intermediary between domestic ducks and humans, as the transmission of mammalian virus strains directly to domestic poultry is unlikely to be a factor in the generation of new pandemic strains. Aquatic birds migrating or overwintering in the region might provide a source of virus for domestic ducks. Yasuda et al. (1991) showed that domestic ducks harbour H3 influenza viruses antigenically and genetically similar to those in pigs, suggesting they may play a role in the transfer of avian influenza viruses from feral ducks to pigs.

Influenza viruses infect a large variety of animals, and species barriers are less important in their ecology than they were thought to be. Given the worldwide interaction between humans, pigs, birds, and other mammalian species there is a high potential for cross-species transmission of influenza viruses in nature. A proposed model of the animal reservoir of influenza A viruses is shown in Fig. 30.2.

The ability of an influenza virus to cross between species is controlled by the viral genes, and the prevalence of transmission will depend on the animal species. The theory that pandemic influenza viruses arise as a result of adaptation and/or genetic reassortment requires that viruses pass either from other animals to humans or vice versa, and that genetic reassortment then occurs by dual infection which results in progeny virus with the ability to infect and cause disease in humans, but with antigenic determinants different from recent viruses affecting the human population. It would seem reasonable to suppose that such transference between species would occur many more times than when the conditions are optimal for the emergence of pandemic viruses. While viruses do not pass to and from humans and animals with complete freedom, under some conditions such transmission does occur.

In January 1976 an H1N1 virus, identical to viruses isolated from pigs in the United States, was isolated from a soldier who had died of influenza at Fort Dix, New Jersey, USA. At least five other servicemen were shown by virus isolation to be infected, and serological evidence suggested that some 500 personnel at Fort Dix were, or had been, infected with the same virus (Hodder et al. 1977; Top and Russell 1977). With the 1918 pandemic in mind this stimulated a universal vaccination programme in the USA, that was eventually abandoned when it became clear that the virus had not spread any further.

The Fort Dix incident cannot be regarded as evidence of zoonoses, since it was not established whether the likely source was pigs. However, there is considerable evidence that transmission from pigs to humans does occur. Kluzka et al. (1961) reported that humans working with pigs in Czechoslovakia had antibodies to the H1 subtype from pigs, and Schnurrenberger et al. (1970) reported that people in the USA who had close contact with pigs were more likely to have antibodies to classical swine H1N1 influenza virus than those who did not.

Final confirmation of the zoonotic nature of H1N1 influenza viruses from pigs came in 1976, when clinical influenza appeared in a herd of pigs on a farm in Wisconsin 2–3 days before a caretaker also became ill with influenza. Viruses isolated from the pigs and the man were shown to be both antigenically and genetically identical swine H1N1 influenza viruses (Easterday 1980). There were several reports in 1994 from North America of swine virus being isolated from humans with respiratory illness, occasionally with fatal consequences (Wentworth et al. 1994). All cases examined followed contact with sick pigs and were due to viruses most closely related to classical swine H1N1 influenza virus. In Europe, De Jong et al. (1986) reported the isolation of classical swine H1N1 influenza virus from three unrelated human cases of respiratory illness, one of which involved a 3 year old child who had apparently not been in contact with pigs, although there had been recent epizootics in pigs in the region in which she was living. Perhaps of greater significance for humans is a report of two distinct cases of infection of children in the Netherlands during 1993 with H3N2 viruses, whose genes encoding internal proteins were of avian origin (Claas et al. 1994). Genetically and antigenically related viruses had been detected in European pigs (Castrucci et al. 1993), raising the possibility of potential transmission of avian influenza virus genes to humans following genetic re-assortment in pigs.

H3N2 influenza viruses are ubiquitous in animals and endemic in most pig populations worldwide. There is no apparent evidence of pigs being infected with this subtype prior to the pandemic in humans in 1968. Indeed, the appearance of an H3N2 subtype variant strain in the pig population of a country appears to coincide with the epidemic strain infecting the human population at that time (Brown et al. 1995b).

Further evidence of the spread of influenza viruses from humans to pigs was the appearance in pigs of H1N1 viruses (or antibodies to H1N1) related to those circulating in the human population since 1977 (Nerome et al. 1982; Brown et al. 1995b). Genetic analysis of two strains of H1N1 virus isolated from pigs in Japan revealed that the HA and NA genes were most closely related to those of human H1N1 viruses circulating in the human population at that time. In addition, re-assortant viruses with some characteristics of human H1 viruses were isolated from pigs in England (Brown et al. 1995a).

Finally, in 2009, the triple re-assortant pandemic H1N1 virus appeared in Mexico and South California.

In historical accounts of pandemics in humans, frequent reference is made to similar diseases in horses occurring either simultaneously or preceding that in humans. Beveridge (1977) noted such references in the accounts of 12 pandemics occurring during the eighteenth and nineteenth centuries.

Serological studies have revealed the presence of antibodies to equine H3 (equine-2) viruses in the sera of people born in the nineteenth century, and this has been considered as possible evidence that virus of this subtype was responsible for the pandemic of 1889–90. Experimental infection of human volunteers with H3 (equine-2) viruses has produced an influenza-like illness with virus shedding and seroconversion. There is no evidence of infection of humans with the other subtype of influenza, H7 (equine-1), which has caused widespread epizootics in horses.

There have been several isolated reports of infection of horses with subtypes H1N1, H2N2, and H3N2, usually associated with human infections. Experimental infections of horses with human-derived H3N2 virus has confirmed their susceptibility to this virus (Blaskovic et al. 1969).

Although there is convincing evidence that all 16 subtypes of influenza A viruses are perpetuated in the aquatic bird populations of the world, only a few of the numerous subtypes have been observed in non-avian hosts. Phylogenetic analyses have revealed that some human pandemic strains most probably emerge following re-assortment between avian and human influenza viruses, with the pig being favoured as the possible intermediate host.

The 1957 pandemic virus had the HA, NA and PB1 genes from an AI virus, while the 1968 H3 virus had two AI virus genes, the HA and PB1. Avian H3N2 viruses are readily transmitted to pigs in south east Asia.

Outbreaks of influenza in pigs in Europe since 1980 have been associated with influenza A viruses which are antigenically and genetically distinguishable from classical swine H1N1 viruses but closely related to H1N1 viruses isolated from ducks. All of the gene segments of the prototype viruses were considered to be typical of viruses of avian origin, indicating that transmission of a whole avian virus into pigs had occurred. These viruses circulated in European pigs and were reintroduced to turkeys causing economic losses.

In 1979, around the Cape Cod peninsula in North America, high mortality in the population of harbour seals was attributed to an H7N7 influenza virus which was isolated from the lungs and brains of dead seals. Antigenic and genetic analyses revealed that the virus was most closely related to viruses from avian species. During the initial studies, four people involved in postmortem examinations of the seals had developed purulent conjunctivitis within 2 days of known contamination with seal material. Although no virological studies were carried on these cases, in subsequent laboratory studies an infected seal, known to be shedding virus, sneezed directly into the eye of one of the investigators who developed conjunctivitis within 2 days. Virus identical to the seal virus was isolated from the affected eye for 4 days after this incident (Webster et al. 1981).

In March 1989 a severe outbreak of respiratory disease in horses occurred in China. An influenza virus of subtype H3N8 was isolated and was antigenically and genetically distinguishable from equine-2 (H3N8) viruses, being most closely related to avian H3N8 influenza viruses. Genetic evidence suggested that this virus was transmitted to horses without reassortment.

Transmission of avian influenza viruses to other species of mammal such as whales and mink have been reported. Genetic analysis of the viruses from whales confirmed that they had probably been introduced from birds. The potential susceptibility of mink to avian influenza viruses had been demonstrated following experimental infections.

Viruses antigenically identical to human variants of H3N2, H2N2, and H1N1 subtypes have been isolated from wild birds and domestic poultry, some have been reported to cause disease outbreaks in chickens that have shown a temporal relationship to influenza epidemics in humans.

Classical swine influenza viruses have also been isolated from ducks (Butterfield et al. 1978; Hinshaw et al. 1978), providing further supportive evidence for the natural transmission of influenza A viruses between avian and mammalian species.

There have been a number of sporadic infections with LPAI viruses in humans. For example, in 1996, a 43 year old woman in England developed self-limiting conjunctivitis. An H7N7 LPAI virus was isolated from an eye swab. She had been looking after domesticated ducks of various breeds that mixed freely with wild ducks on a small lake. An H9N2 LPAI virus was isolated from 7 children and adults in mainland China and Hong Kong in late 1998 and early 1999. The two patients in Hong Kong had developed fever and a chest infection, all recovered. In 2002, there was an H7N2 LPAI outbreak in the USA. A patient with serious underlying medical conditions developed respiratory symptoms and an H7N2 LPAI virus was isolated. Finally, an H7N3 HPAI virus was detected in samples from 2 Canadian adults with conjunctivitis, coryza, and headache.

In contrast, most human AI infections with fatal outcomes involved H5N1 infections in Asia and H7N7 HPAI viruses in the Netherlands.

Finally, in the early 2000s, an H9N2 LPAI virus infected 8 people, that appeared to be widespread in other parts of Asia (Choi et al. 2004).

The ability of an influenza virus to spread is related to both the virus and the host involved. The age, population density, and air space may all affect transmissibility. Host-specific or host-adapted viruses have higher affinity for spread within a given host, than after transmission to a ‘foreign’ host. For example, the human viruses which cause epidemics and pandemics have a high capacity for spread amongst susceptible individuals. In contrast, the transfer of non-re-assorted influenza viruses derived from birds or pigs, produce only mild or inapparent infection in humans and rarely result in secondary transmission. The majority of infections of humans in the USA with classical swine H1N1 viruses have not resulted in transmission from infected to in-contact individuals (Dasco et al. 1984; Patriarca et al. 1984).

The success of interspecies transmission of influenza viruses depends on the viral gene constitution. Successful transmission between species can follow genetic reassortment, with a progeny virus containing a specific gene constitution having the ability to replicate in the new host. Re-assorted viruses may have a relatively low fitness, and will not be able to perpetuate in the new host. These observations support the potential role of the pig as a mixing vessel of influenza viruses from avian and human sources. The pig appears to have a broader host range in the compatibility of the NP gene in re-assortant viruses than both humans and birds. Furthermore, cell receptors for both avian and human viruses are present in the pig trachea and the intermediate temperature of pigs compared to humans and birds may be important since virus synthesis is influenced by temperature control.

Transmission of the H5N1 viruses to, and between, humans seems limited and that could be due to incomplete host adaptation or a dose response restriction.

In addition, close contact is required to infect other humans. The main concern is that a seasonal influenza A virus will infect an H5N1-infected human or other mammal and a reassortant virus containing the correct combination of genes for efficient human-to-human transmission will emerge.

Generally, two control measures are available for influenza in humans: immunoprophylaxis with vaccines, especially in older adults (Monto et al. 2009), and chemoprophylaxis or therapy with antiviral drugs.

Since the late 1940s, the principal preventive measures against influenza have been inactivated virus vaccines. The efficacy of vaccines has varied between 60–90% and has been dependent on the closeness of the ‘antigenic match’ between the vaccine virus and the epidemic virus. However, even for those not completely protected, vaccination reduces the severity of the disease, thereby reducing costs and mortality (Fiore et al. 2009). Two basic approaches for immunization have been pursued: the use of inactivated virus preparations and the use of live, attenuated viruses.

Influenza vaccine is prepared from purified, embryonated egg grown viruses that have been rendered non-infectious. Four major innovations have been incorporated: the use of zonal centrifugation, the use of ether or other lipid solvents to disrupt the virus, the introduction of high-yield reassortants to improve yields in the chick embryo, and the development of better methods to quantitate the amount of viral antigens present in the vaccines. All of these efforts have led to inactivated vaccines that are better purified and more predictable in their reactogenicity and immunogenicity.

Each year the influenza vaccine is redefined to reflect changes in the antigenicity of circulating virus strains and contains virus strains representing influenza viruses believed likely to circulate in the forthcoming ‘influenza season’. At present this involves two type A viruses, H1N1 and H3N2, and one type B virus (WHO). Recommendations for influenza vaccine composition. (Available at: www.who.int/csr/disease/influenza/vaccinerecommendations1/en/index.html.)

The exact strains of these viruses to be used are identified by an international network of laboratories that maintain surveillance for new influenza virus variants throughout the world. These laboratories are coordinated through the WHO.

The composition of the vaccine rarely causes systemic or febrile reactions. Whole virus, subvirion, and purified surface antigens are available. Depending on the age group, the response to inactivated vaccines is either a primary or booster type of immune response. Children who have not been exposed to influenza mount a primary antibody responce, and titres after the first dose of vaccine are low. After a second dose of vaccine which provides a boost, antibody titres rise in these children. Most adults, unless they are being exposed to an entirely new antigen, will mount a booster antibody response, even to strains whose antigens are marginally different. Antibodies to influenza vaccines are generally of the IgG subclass, and they react against the HA and NA of the vaccine strains. Antibody titres usually peak at 10–14 days after vaccine boost, then decline in the ensuing months. Serum antibodies appear to be very important in protecting against infection with influenza viruses. Extensive data show that the serum HI antibody titre correlates inversely with the occurrence of established infection with influenza viruses.

Most vaccinated children and young adults develop high post-vaccination HI antibody titres. These titres are protective against infection by strains similar to those in the vaccine or the related variants that emerge during outbreak periods. Elderly people and those with certain chronic diseases may develop lower post-vaccination antibody titres than healthy young adults and thus remain susceptible to influenza upper respiratory tract infection. Nevertheless, even if such people develop influenza illness, the vaccine has been shown to be effective in preventing lower respiratory tract involvement or other complications, thereby reducing the risk of hospitalization and death.

The effectiveness of influenza vaccine in preventing or attenuating illness varies, depending primarily on the age and immunocompetence of the vaccine recipient and the degree of antigenic similarity between the virus strains included in the vaccine and those circulating during the influenza season. When there is a good match between vaccine and circulating viruses, influenza vaccine has been shown to prevent illness in approximately 70% of healthy children and young adults, while preventing hospitalization for pneumonia and influenza among elderly people living in the community. Adverse reactions to vaccination can occur locally at the vaccination site, including pain and erythema, dependent on the age group, vaccine type, and route of inoculation. Fever and systemic signs occur less frequently, in the range of 5–30%, and allergic reactions have been noted rarely, presumably to a vaccine component such as egg protein.

A live attenuated, cold-adapted (ca) re-assortant influenza virus vaccine was licensed in the USA in 2003. This vaccine relies on the use of an attenuated donor virus to confer the property of attenuation to contemporary wild-type strains by genetic re-assortment. This vaccine is very stable, owing partly to the multigenic requirement for the attenuated phenotype, and is at least as good as the inactivated vaccine in a population previously exposed to influenza virus (Fiore et al. 2009).

The issue with inactivated vaccines include limited global influenza vaccine manufacturing capacity, up to 6 months production time, require large supplies of chicken eggs and biological containment facilities. Plasmid (Deoxyribonucleic acid) DNA vaccines can be produced rapidly using simple bacterial fermentation procedures. DNA vaccines are a potentially powerful approach to the development of subunit vaccines (Robinson et al. 1993; Ulmer et al. 1993; Smith et al. 2010). Vaccination by DNA inoculation is achieved by the uptake and expression of the inoculated DNA. The protein that is expressed by host cells raises the immune response including inducing antibody and cytotoxic T-cell responses.

The antiviral drugs mentioned previously are effective when used as prophylaxis.

Hong Kong poultry in Hong Kong SAR are free of H5N1 due to the extensive culling programme together with a well managed surveillance programme and successful vaccination approaches (Ellis et al. 2004). The control programme aim is to prevent, manage or eradicate AI. In order to be effective, controls have to include biosecurity aspects to prevent the introduction or escape of the pathogen at the farm level, maintain rapid, sensitive and specific diagnostics and surveillance, managing acutely infected or convalescing animals, decreasing host susceptibility to the pathogen by immunization and educating poultry farmers and market stall holders in order to prevent transmission or spread. Increased cooperation between the veterinary and human health agencies within countries will ensure control of AI viruses and manage the public health implications.

Vaccination of people at high risk before each annual influenza season is currently the most effective measure for reducing the impact of human influenza. When vaccine and epidemic strains of virus are well matched, achieving high vaccination rates among closed populations can reduce the risk of outbreaks by inducing ‘herd’ immunity. This occurs when the overall number of susceptible people in a population becomes too low for virus to spread and infect a significant number of the susceptible individuals.

To maximize protection of persons at high risk, they and their close contacts should be targeted for organized vaccination programmes. Influenza vaccination is strongly recommended for any person above 6 months of age who, because of age or an underlying medical condition, is at increased risk for complications of influenza. The high-risk group consists of: persons above 65 years of age, residents of nursing homes and other chronic-care facilities, people with chronic disorders of the respiratory or cardiovascular systems, and people who have suffered chronic metabolic diseases or immunosuppression (including as a result of medication) within the past year. In addition, people who may transmit influenza to those at high risk, i.e. hospital and nursing home personnel and household members of people in high-risk groups, should also be vaccinated. Other groups may be included on individual merit, such as pregnant women, people infected with human immunodeficiency virus, and foreign travellers.

Although an influenza vaccine may contain one or more of the antigens administered in previous years, annual vaccination using the current vaccine is necessary because immunity declines within the year following vaccination. Old batches of vaccine should not be administered as the constituent virus strains are updated annually to reflect the predicted epidemic strains for the coming year.

Beginning each September, when vaccine becomes available for the forthcoming influenza season, people at high risk should be offered vaccine. Because influenza activity usually peaks between late December and early March in the northern hemisphere, people in this part of the world should be vaccinated by mid November. However, it is important to avoid administering the vaccine too far in advance of the influenza season in such places as nursing homes because antibody levels may begin to decline within a few months of vaccination. Earlier vaccination is warranted, however, in particular situations, such as the early onset of an epidemic. Unvaccinated children need two doses of vaccine, at least a month apart, and, under normal circumstances, the second dose should be given before December.

Each year influenza viruses isolated from epidemics are characterized antigenically in WHO influenza reference laboratories, and this information is used to evaluate the antigenic similarity with the virus strains incorporated into the current vaccine. This will provide some information on the potential efficacy of the vaccine, since the strains included in the vaccine had been selected ahead of the new ‘influenza season’.

Avian influenza is primarily a disease of birds, however, certain subtypes, notably H7 and H5, can be transmitted directly to humans. Although this is rare it can cause illness in humans and H5N1 infection, in particular, is associated with a high case fatality rate. Avian viruses can be categorized into high and low pathogenicity strains, which relates to their virulence in birds. There is no evidence that virulence correlates in poultry are the same in humans. During the low path H7N2 avian influenza virus outbreak in the UK in 2008, three of the four clinical cases were admitted to hospital for a period. Outbreaks are more common than previously recognized. Table 30.4 summarizes recent avian outbreaks in UK.

There is also a risk that co-infection with an avian influenza virus and a human influenza virus could give rise, through genetic reassortment, to a hybrid virus to which there was little or no population immunity, that is capable of causing human illness and would be able spread from person to person effectively. Such a virus would meet all the prerequisites of a virus capable of causing a global pandemic. As such an important aspect of the global monitoring of influenza is detecting and investigating human disease caused by avian influenza viruses.

There are therefore three main policy aims relating to the human health aspects of avian influenza incidents:

Protecting the health of those involved in an incidental exposed to the virus

EU legislation requires the rapid implementation of containment measures in affected premises when an outbreak of a highly pathogenic avian influenza, of any type, occurs (Council Directive 2005/94/EC).

These measures include movement restrictions within a prescribed area and the culling of the birds on the premises. Movement restrictions mean that there will be animal welfare issues involving contact between workers and infected birds and their bedding or faeces. The culling of the birds and the disposal of the carcasses, bedding and faecal material will also involve close contact with potentially infected material.

In 2003, a fatal human case of influenza A/H7N7 infection occurred in a veterinarian in the Netherlands involved in controlling an outbreak on a farm. The patient, and various workers and their families who became infected, had not taken measures to protect themselves from infection. This incident demonstrated the necessity of stringent protective measures for people who have close contact with sick and recently dead birds affected by any highly pathogenic avian influenza (Koopmans et al. 2004).

The transmission risk of low pathogenicity AI viruses to poultry handlers is the same as the risk of high pathogenicity virus transmission and that there is no evidence that virulence correlates in poultry are the same in humans.

In order to minimize exposure to the influenza virus, workers involved in caring for the birds, the culling of the birds, and the disposal of the carcases, the bedding and faeces, are required to wear personal protective equipment (PPE). This involves the wearing of protective overalls, gloves, boots, respirators and eye protection when there is the potential for contact with the birds, their bedding or faeces. Guidance on the recommendations for the use of PPE has been developed by the European Centre for Disease Prevention and Control (ECDC).

Since the use of PPE is highly dependent on individual compliance and in acute situations it may be used by relatively inexperienced individuals, there is still a significant risk of exposure to the virus therefore prophylaxis with an antiviral is given. The neuraminidase inhibitor oseltamivir is usually used and this is given as a daily dose of 75mg for the duration of exposure and continued for 10 days after the date of last exposure. Prophylaxis should be started before exposure occurs. All individuals coming into contact with the birds, their bedding or faeces also have their health monitored and any illness is quickly investigated. The value of this monitoring was demonstrated during an outbreak of a low pathogenicity H7N3 avian influenza virus in the UK in 2006 when a conjunctivitis caused by the virus was detected in a worker (Nguyen-Van-Tam et al. 2006).

In addition to these measures those exposed prior to the infection being detected, for example workers on the facility, are offered post-exposure prophylaxis–oseltamivir 75mg for 10 days and undergo health monitoring while taking prophylaxis.

Protecting the health of public from exposure to food stuffs that may be contaminated The risk to the general public from avian influenza in general and specifically from consuming meat and poultry products has been repeatedly considered to be minimal by authorities such as the ECDC, the European Food Safety Authority (EFSA), national food safety agencies within the EU and the World Health Organization International Food Safety Authorities Network.

Guidance focuses on the importance of cooking and food hygiene but this is mostly to protect against other pathogens (Campylobacter, Salmonella and E. Coli O157) that are more common, better adapted to humans and much more infectious.

The possibility of transfer of an avian influenza in or on unprocessed food or fomites such as packing or clothing and the potential for this to pose a threat to workers in the food or catering industry and those preparing food in the home is believed to be very low, as the virus remains poorly adapted to humans.

Minimizing the risks of a new influenza subtype emerging

A major concern around the handling of avian influenza incidents is the possibility of the co-infection of a worker/responder with the outbreak avian strain and a seasonal human influenza. It is theorized that such an event could lead to the creation of a new influenza subtype through genetic re-assortment which could become a pandemic strain. Low pathogenicity in poultry does not indicate a reduced tendency to transmit to humans and therefore the transmission risk of low pathogenicity AI viruses to poultry handlers is the same as the risk of high pathogenicity virus transmission.

Vaccination with seasonal influenza vaccine is therefore recommended for poultry workers, cullers, veterinarians and others who could come into contact with potentially infected birds. In the UK this operationalized by inclusion within the national influenza vaccine programme.

Vaccine is also offered to those unimmunized workers who are involved in an avian influenza response however, given that it takes between seven to ten days to mount an adequate antibody response it is difficult to see the justification for this. Proactive vaccination within the groups likely to be involved would appear to be a better strategy. It is therefore important to emphasize that other public health measures in managing an acute incident are probably of greater importance. These include preventing unnecessary exposure, use of PPE, use of Neuraminidase Inhibitors (NI) prophylaxis, and health status follow up.