The present outbreak of avian influenza in Asia is the largest ever described. Within a few months, more than 100 million birds have either died or been culled. Even though the infection could be transmitted to humans with resulting severe disease or death, the virus is not well adapted to man, and there is no evidence of human-to-human transmission. The greatest threat to public health therefore lies in the risk of either reassortment of virus (combination of avian and human influenza), or adaptation of the influenza virus to humans, resulting in a new pandemic. The present situation in Asia stresses the importance of having updated pandemic plans in all countries.
Influenza in birds
Birds, especially sea and water birds (such as ducks and geese), are the natural hosts of influenza A. The influenza A viruses are divided in subtypes according to their surface glycoproteins; hemagglutinin (H) and neuraminidase (N). There are 15 different hemagglutinins and nine different neuraminidases in influenza A, all of which could cause avian influenza. In all, more than 80 different avian influenza subtypes have been described.
The majority of all avian influenza viruses give rise to a mild respiratory tract infection, or no disease at all in birds. However, the virus could mutate into an immensely more virulent variant, with a case fatality rate among birds exceeding 50 %, sometimes approaching 100 %. This aggressive variant was first described in 1878, under the name of “bird plague” (1). The modern name is highly pathogenic avian influenza (HPAI).
This far, only two H types (H5 and H7) have been shown to cause HPAI, as the virus has changed from a relatively harmless to a highly pathogenic one (Table 1). The hemagglutinin glycoprotein is produced as a precursor (HA0), which requires post-translational cleavage by trypsine-like proteases of the host to be functional. The virus could therefore only replicate in those parts of the bird where these enzymes are present, i.e. in the respiratory and gastro-intestinal tracts. In HPAI the viruses have incorporated multiple basic amino acids at the cleavage sites, enabling the viruses to replicate throughout the bird, with rapid damage of vital organs and death (2,3).
Such mutational events have been described in the USA in 1983–1984, when an originally low pathogenic H5N2 avian influenza (LPAI), mutated to become HPAI, and in Italy in 1999–2001, when the same thing happened to a H7N1 virus. The case fatality rate in both outbreaks approached 90 %.
Chickens and turkeys are the species most susceptible to HPAI, but different variants of avian influenza viruses may have different affinity to different birds. A virus that could wipe out flocks of turkeys could totally spare the chickens (4).
Spread of avian influenza
Droppings from migrating water birds could contain large amounts of influenza viruses – enough to spread the disease over large geographical areas. The virus could survive for extended periods of time in low temperatures. Even though these migrating birds are responsible for most onsets of avian influenza outbreaks, further transmission usually take place between flocks of domestic birds through human activities (equipment, trucks, soiled boots, etc) (5). The large Asian food markets with live fowls in abundance are believed to have played a crucial role for many outbreaks. Except for culling all flocks in the vicinity of a known outbreak (often within a radius of three kilometres), immaculate hygienic measures are therefore essential in preventing further spread of disease. Vaccinating flocks in outbreak areas has been another strategy, but vaccination may mask disease in infective birds.
Also contaminated animal carcasses can transmit disease, and if disposals from infected animals reach flocks in other countries or regions, new outbreaks could result. This is the main reason for the bans of import of poultry posed by the European Union in January (6). Similar actions have been taken by a number of other countries. The outbreaks have had enormous economical consequences. In the American and Italian outbreaks mentioned previously, more than 30 million birds were culled. The present Asian outbreaks of influenza A/H5N1 are the most extensive outbreaks ever described, and so far more than 100 million birds have either died from the disease or been culled. Officially recognised outbreaks have been reported from Japan, Korea, Vietnam, Thailand, Indonesia, Cambodia and China, including Hong Kong. Even though the first outbreaks were not officially confirmed until December 2003, the disease may have been present in the region since the spring of 2003. Several outbreaks of lethal disease in fowl during the last year have been attributed to Newcastle disease, another epizootic fowl disease.
Avian influenza in humans
Avian influenza virus is strongly adapted to birds, but may also infect domestic pigs. Other mammalian species appear normally not susceptible. Up until 1977, the species barrier was considered sufficient for preventing direct transmission of avian influenza to humans. Transmission from birds to humans was first evident in Hong Kong in 1997, when an H5N1 strain caused respiratory tract infection in 18 persons, of whom six died (7,8,9). This happened at a time when bird farms of the city were struck by avian influenza, and it could be shown that the virus isolated from humans was identical to the virus circulating in birds (10). The outbreak was halted after a tremendous effort, when all of the 1.5 million chicken in the city were culled within three days. Since then further outbreaks with transmission of avian influenza to humans have been described (Table 2). When the first news of a mystery disease (that was later shown to be SARS) were reported from China, there was an initial relief that this disease was not avian influenza.
The clinical picture of avian influenza in humans in the 1997 Hong Kong H5N1 outbreak has been described in detail (9). Seven of the patients had mild symptoms, including fever, throat pain, conjunctivitis and mild respiratory tract symptoms. The additional 11 patients also had more serious symptoms; pneumonia, hepatopathia, renal failure and pancytopenia. The case fatality rate was lower in children below the age of 12 years (1 of 9) than in patients above the age of 12 (5 of 9). In the Dutch outbreak in 2003, the majority of the 86 patients had only mild disease (mostly conjunctivitis). Seven of the cases also had influenza-like illness, and a veterinarian had severe disease with viral pneumonia and died (11). Notable is that H7N7 virus was detected in three family members of cases, who had themselves not been in contact with infected poultry. This indicates that human-to-human transmission occurred in the Dutch outbreak.
In the present Asian outbreaks, severe disease with a high case fatality rate has been reported from Vietnam and Thailand. As of March 24, 22 cases with 15 deaths have been reported from Vietnam and 12 cases with 8 deaths have been reported from Thailand. Detailed descriptions of the first ten Vietnamese patients (eight of whom died) have recently been published (12). The mean age was 14 years, and only two were older than 20 years of age, six were males. All had cough, dyspnoea, and fever > 38.5°C. Five of the patients had sputum, and seven had diarrhoea. Rash, myalgia or conjunctivitis was not noted in any of the patients. Nine of them had a clear history of direct contact with poultry (three days median time before onset of illness).
The 32 Asian patients have mainly been reported from tertiary care hospitals. All experience from other emerging infectious diseases, indicate that the first reported patients are a sub-selection with especially severe disease. In addition studies on health care use among patients with respiratory symptoms in Vietnam have shown that a large majority of patients only use self-medication or seek care within the unregulated private sector where diagnostic capacity and quality of care is low (13,14). Potential cases of HPAI may thus pass unnoticed to the national health care sector and official reporting system. One could therefore assume that another spectrum of disease would be identified in population based epidemiological and microbiological studies in affected areas.
Disease risks in humans
The only four influenza A subtypes that have been shown to circulate between humans since the Spanish flu epidemic are H1N1, H1N2, H2N2, and H3N2. However, avian influenza viruses (H5N1, H7N7 and H9N2) have been transmitted from birds to humans through direct contact with infected birds or their droppings (Table 2). Persons with occupational exposure to fowls are especially exposed to this risk. Considering the enormous number of infected birds in the last months (counted in tens of millions) and the close proximity in which people and birds are residing in these countries, the individual risk of a person to fall ill with severe avian influenza must be very small. An effective way of further decreasing this risk is to avoid close contact with birds and bird droppings and to adopt strict personal hygiene with frequent hand washing (15). The H5N1 virus is poorly adapted to humans, and to this day there has been no evidence of transmission between humans. There is therefore no reason to issue travel warnings to any affected country, as was done during the SARS epidemic last year.
There are also no data indicating transmission to humans through infected meat or eggs from diseased birds (16). Import restrictions on poultry from affected countries are only posed to prevent outbreaks in fowls and not of a public health concern. Deep-freezing of poultry meat does not affect the virus, but heating to 70° is an effective way of killing the influenza virus. This is of course recommended regardless of any avian influenza outbreak.
Vaccination and treatment
There is presently no vaccine against H5N1 influenza in humans, but preparations for such production are on the way within the WHO network for influenza surveillance. However, these vaccines will not be on the market until the autumn of 2004. Avian influenza virus is susceptible to the modern neuraminidase inhibitors (oseltamivir and zanamivir). Data on a rapid development of resistance are from two older drugs (amantadin and rimantadin).
Risks of a new influenza pandemic in humans
Influenza A virus lacks mechanisms for proofreading and correction of mutations during the replication, and is therefore genetically unstable. Under the pressure of constant small changes in the genome, new antigenic variants appear on a regular basis in a process called “antigenic drift”. Two different virus subtypes can also combine to a new virus – reassortment (17). The result may be a compeletly new virus variant – an “antigenic shift”. Pigs are susceptible to both human and avian influenza, and may act as “mixing vessels”, and an important mediator in the genesis of new viruses. If sufficiently contagious and virulent, such virus could spread very rapidly, causing severe disease. This is how two of the global influenza pandemics of the last century are thought to have started; the Asian flu in 1957–58 (influenza A/H2N2) and the Hong Kong influenza in 1968–69 (influenza A/H3N2). The origin of the virus causing the Spanish flu in 1918–1920, with 45 million deaths, is still under debate.
In the present unstable situation in Asia, the greatest public health threat is therefore not the avian influenza itself, but rather the inherent risk of a recombination of H5N1 virus and a human influenza virus, with the potential of starting a new pandemic. The strategy of the WHO rests on an active international surveillance aiming at early identification of such a virus, and a subsequent fast manufacturing of effective influenza vaccines. The present situation should also be a clear signal for all nations to have updated national influenza pandemic plans.