Policy ForumPUBLIC HEALTH

Surveillance of Animal Influenza for Pandemic Preparedness

See allHide authors and affiliations

Science  09 Mar 2012:
Vol. 335, Issue 6073, pp. 1173-1174
DOI: 10.1126/science.1219936

The 2009 H1N1 pandemic was not as severe as initially feared. This has led to complacency in some quarters that future pandemics will be of comparable impact and as readily dealt with. However, by September 2009, just 5 months after the recognition of the novel pandemic H1N1 virus, almost 50% of children in Hong Kong were already infected (1), which reflects the speed of spread of the virus to and within international travel hubs. In most parts of the world, vaccines were not available in time to substantially affect the first wave of disease. A more virulent virus, such as one comparable to the 1918 H1N1 virus or the H5N1 “bird flu,” spreading with such speed would be a global catastrophe.

The 2009 pandemic virus was circulating in humans in Mexico for about 4 months before it was recognized (2). The first known cases of severe acute respiratory syndrome (SARS) occurred in Foshan City, Guangdong, China, in mid-November 2002, and triggered initial investigations by the Guangdong Health Bureau in mid-January 2003; the etiological agent was identified in mid-March 2003 (3). Earlier detection of novel pandemic viruses once they have emerged in humans would provide longer lead times for vaccine development and is crucially important. Pandemic influenza viruses arise, in part or whole, from influenza viruses of other animals. Precursors of the 1957 and 1968 pandemics circulated in animals for some years before emerging in humans (4). What is the role of influenza virus surveillance in animals for pandemic preparedness?

Some criteria that can be used to assess the risk of animal influenza viruses include the high prevalence of animal influenza viruses in species found in close proximity to humans (e.g., domestic livestock); those undergoing rapid genetic change or reassortment; viruses with a predilection for binding to α 2,6-linked sialic acid (the influenza receptors found on human upper respiratory epithelia) (57); efficient infection of ex vivo organ cultures of the human upper respiratory tract (nasopharynx and bronchus) (8, 9); and transmission by airborne droplets in ferrets, the best available animal model for human transmission (10, 11). Evidence of repeated zoonotic transmission to humans and lack of cross-reactive “herd-immunity” in the human population are also important. The severity of disease associated with zoonotic human infection should also be considered, not because severity alone makes a virus more likely to be pandemic but because preemptive preparations are more rationally targeted at viruses with potential to cause exceptionally severe pandemics.

Avian and swine influenza virus statistics.

(Top) The number of avian and swine influenza viral genome sequences (by year of deposit) available in the Influenza Viral Sequence Database (29). Only nonidentical viral nucleotide sequences were counted. (Bottom) Cumulative sequence data by continent in which sampling was done.

Identification of molecular signatures associated with transmissibility of animal viruses to humans and between humans will be an important additional parameter. Attempts to define molecular markers associated with host adaptation and transmission have been under way for many years but have gathered speed with development and integrated application of virus reverse genetics, glycan arrays, studies in ferrets, and the use of ex vivo organ cultures of the human respiratory tract. Recent studies have tried to elucidate the molecular changes that allowed the 1918 H1N1, 1957 H2N2, and 2009 swine-origin pandemic H1N1 viruses to become transmissible in humans (7, 8, 1214). Investigations of changes that confer mammalian transmissibility to avian H9N2 (11), H7 (6), and H5N1 viruses (10, 15) have begun to provide insights. Future studies will try to identify common mechanisms across avian influenza virus subtypes that confer increased mammalian transmissibility.

Before 2009, the application of ecological, virological, host-immunity, and molecular criteria led to the identification of virus subtypes H2, H7, H5N1, and H9N2 as potential pandemic threats (16). These criteria would also have led to identification of the precursors of the 2009 pandemic, which were undergoing rapid genetic reassortment (17), with receptor binding to α 2,6-linked sialic acids and with evidence of repeated human transmission (18). The reason why they were not high on the list of pandemic candidates was the (mistaken) belief that a pandemic could not arise from viruses with hemagglutinin subtypes already endemic in humans.

Virological and molecular surveillance of influenza viruses in animals is ongoing (see the graphs). In practice, there are a number of constraints. Reports of outbreaks may lead to culling of farm animals and to restrictions in international trade, which could lead to economic losses to the farmer and the country. Furthermore, whereas veterinarians focus on viruses with economic impact on livestock, viruses of pandemic risk may not necessarily cause illness in the animal reservoir. Indeed, this was probably the case with the precursors of the 1957, 1968, and 2009 pandemics. Thus, surveillance needs to be targeted at healthy as well as ill animals, an effort that is unlikely to be of high priority for animal health. Detecting influenza viruses in apparently healthy domestic livestock may make them unmarketable, which explains the reluctance in the swine industry of developed, as well as developing, countries to permit surveillance of their swine herds for influenza viruses.

However, appropriate choice of strategy can help address some of these concerns. For example, the systematic surveillance in live poultry markets in Hong Kong and China (19) provided advance warning (unfortunately not widely heeded) of the impending emergence of highly pathogenic avian influenza (HPAI) H5N1 that caused outbreaks in poultry and disease in humans in many Asian countries in 2004. These warnings prompted Hong Kong to introduce evidence-based interventions in the poultry industry, minimizing the impact of H5N1 within Hong Kong (20). Surveillance of pigs in an abattoir in Hong Kong (anonymized and unlinked to supplying farms) allowed systematic and long-term surveillance to be carried out on swine influenza in southern China (21).

The 2009 pandemic virus, which very likely emerged from swine, spread worldwide in the human population and transmitted from humans back to swine in many parts of the world. Pandemic H1N1 infection in swine has readily reassorted with other swine influenza viruses, to give rise to a range of reassortants (22, 23), some of which have infected humans (24). The 2009 pandemic H1N1 virus is changing more rapidly in swine than in humans (2224), which makes it particularly important to have global surveillance of influenza in swine. However, we still have little or no data on influenza viruses circulating in swine in Mexico, the putative birthplace of the pandemic virus of 2009, or from swine in South and Central America, Africa, and South Asia.

Although current data on influenza viruses in animals are patchy and far from satisfactory, the situation has improved over the last decade (see the graphs). Concern about the spread of HPAI H5N1 has mobilized resources and the “One Flu for One Health” paradigm, which views human, animal, and wild-life health issues in a holistic manner (25), may have also increased awareness. OFFLU is a collaboration among the laboratories of the animal health sector [Food and Agricultural Organization of the United Nations (FAO), and World Organization for Animal Health (OIE)] to improve understanding of the ecology and evolution of animal influenza viruses for assessment of human health risk (26). Capacity-building efforts of FAO, OIE, World Health Organization (WHO), and U.S. Centers for Disease Control and Prevention (CDC); research networks such as those of the Centers of Excellence for Influenza Research and Surveillance (CEIRS) funded by the National Institute of Allergy and Infectious Diseases [NIAID, U.S. National Institutes of Health (NIH)]; and others have contributed to new data. Using H5N1 as an example, virus isolation and sequence data from human and avian isolates are available from many affected countries (supporting online table), through concerted efforts within countries, sometimes assisted by WHO or OIE reference laboratories or through other collaborative arrangements. Although concerns about sharing viruses and benefits associated with surveillance have surfaced in some instances, recent WHO Pandemic Influenza Preparedness (PIP) initiatives have aimed at equity (27). The biannual WHO influenza vaccine strain selection meetings of the Global Influenza Surveillance and Response System, although primarily focused on selection of the most appropriate influenza virus strains for seasonal vaccines, also reports on antigenic and genetic characteristics of zoonotic influenza viruses for pandemic preparedness. Recent reports include data in “real-time” on contemporary virus genetic sequence and antigenic data from many countries (e.g., Vietnam, Egypt, Bangladesh, India, China, and Cambodia) (28).

As we continue to urge more data collection, we should not ignore the progress that has been made and the data that are already available. If molecular signatures for transmissibility of animal influenza viruses in humans are better defined, identifying such mutations in viruses isolated during surveillance in animals (and from zoonotic transmission events in humans) might be possible and would be a further incentive to enhance animal surveillance.

The research questions pertinent to influenza are also relevant to the broader challenges posed by other emerging infections (e.g., SARS and Nipah virus). There is no guarantee of short-term success. However, understanding the viral and host determinants that permit animal influenza and other emerging viruses to transmit to and between humans is arguably one of the most important research questions today. Discussions have started (and need to continue) on how the various types of data may be used for risk assessment and pandemic preparedness and on what concrete actions may follow. Even though a specific pandemic precursor may or may not be identified in time to allow preemptive intervention, prioritizing influenza virus subtypes and antigenic variants for vaccine preparation will be a major step forward. The long-term benefits of such research, which are less readily communicated to the lay public than fears, should not be ignored. Nature remains the most efficient bioterrorist of all!

References and Notes

  1. We thank H. Yen for critical comments on the manuscript. Supported by the Area of Excellence Scheme of the University Grants Committee (AoE/M-12/06), Hong Kong SAR, and contract HHSN266200700005C from NIAID, NIH, USA.

Navigate This Article