Pandemic influenza awareness week. Day 2: Our adventures with avian flu

Anyone working in the area of influenza virus epidemiology is familiar with the name Robert Webster. A virologist at St. Jude's Children's Hospital in Memphis, the native New Zealander has been leading the charge against influenza for well over 40 years. Barely out of graduate school, Webster hypothesized that something like genetic reassortment (which had not yet been discovered) occurred to cause the big changes that appeared among human influenza viruses, driving pandemics. He performed a simple experiment that cemented the course of his career: he found that serum from patients who had survived the 1957 influenza pandemic reacted with avian influenza viruses. Later genetic analyses showed that the "Asian flu" virus had indeed received 3 of its 8 gene segments from birds. It happened again in 1968: the pandemic virus was the result of a reassortment between human and avian influenza viruses. These observations led to more than 30 years of surveillance of waterfowl in many different countries, and the revelation that these waterfowl constitute a reservoir of all known subtypes of influenza virus.

Webster's worst fears seemed to be coming true in 1997. Hong Kong was experiencing an influenza outbreak in chickens so severe it had been nicknamed "chicken Ebola." Humans were also affected. The first case was in a 3-year-old boy from Hong Kong. Though doctors knew he had died of the flu, they were uncertain of the strain, and sent samples off to several high-level laboratories for further testing. When it came back H5N1, the Centers for Disease Control and Prevention sent Keiji Fukuda to Hong Kong to investigate. After a month of searching, he and his team found no further evidence of infection with this avian virus in the human population--so they left, writing off the boy's illness as a "freak occurrence." They were premature. By the end of the year, 18 cases had been confirmed; 6 died. Clinical features often included a primary viral pneumonia and death quickly after onset of symptoms. The route of transmission in all cases appeared to be direct bird-to-human contact. Fearing a public health crisis, officials ordered the culling of Hong Kong's entire poultry population. Analysis of the virus showed it to be a serotype H5N1 virus.

Though H5N1 has gotten the lion's share of the spotlight, other avian viruses have caused human disease in the past decade. In The Netherlands, an H7N7 virus caused 89 infections with 1 death in 2003. Human infections with avian H7N3 were reported in British Columbia in 2004, and with H7N1 and H7N3 in Italy between 1999 and 2003. Avian H9N2 viruses were in the spotlight in 1999, when they were found in 3 children in Hong Kong who were suffering from flu-like symptoms. All three recovered.

The United States isn't isolated from this type of outbreak, either. During 1983 in Pennsylvania, influenza burned through the poultry population, where health officials there resorted to the same drastic measure they took in Hong Kong: destruction of the entire poultry stock. In 2002, avian H7N2 caused an outbreak among poultry in Virginia, and at least one serologically-confirmed human case. This virus was seen again in the United States in 2003, when a patient showed up at a New York hospital with respiratory symptoms. It's still unknown how he contracted the virus.

Clearly, avian viruses infect humans periodically--quite likely, much more often than we know about. However, H5N1 seems to be different. It has infected, and killed, an incredibly diverse group of animals, including those who normally are unaffected by influenza: blue pheasants, black swans, turtledoves, leopards, mice, domestic cats, even tigers. And both laboratory and clinical analyses conclude that it seems to have gotten worse since 1997.

H5N1 was back on the radar in 2003, when it caused 35 cases (24 of them fatal) in Vietnam and Thailand. In 2004, the virus caused 44 known cases in humans; 32 of them were fatal. Most of these cases were in previously-healthy children and adults. Poultry was again infected, and this time around seemed to be even more virulent. 120 million birds were killed between January and March of 2004. Additionally, pigs had been infected. This was particularly worrisome, as pigs had long been thought to act as a "mixing vessel" between human and avian strains. (Pig cells are able to bind both human and avian hemagglutinin; thus, viruses of both types can replicate within their cells, with the possibility of recombination and spread of a "humanized" avian virus). Animals normally resistant to influenza disease were again affected. Leopards and tigers died in a Thai zoo after consuming infected raw chicken. In the latter case, it seems that tiger-to-tiger transmission occurred as well.

There is also evidence for at least one case of human-to-human transmission during this outbreak. In Thailand, a young girl died, followed shortly by her mother, who had served as her nursemaid. No contact between the mother and poultry could be found. Additionally, an aunt of the girl, who had also assisted in her care, complained of a sore throat, cough, and fever. She was later found to be infected with H5N1 as well. It was feared that this could be the beginning of The Big One. However, no other cases stemmed from these, and it seems likely that the transmission was due to the large amount of time spent in close contact with the girl and her secretions; the virus did not appear to be easily passed person to person--so far.

As of October 2005, cases of avian H5N1 have been reported in Thailand, Cambodia, Indonesia, and Vietnam. Combined, there have been 72 cases and 28 deaths in these countries since the end of 2004, and a total of 116 confirmed cases and 60 deaths since 2003. These are likely underestimates. If anything can be said about H5N1, it's that it is unpredictable. A case was written up in the February 2005 New England Journal of Medicine detailing a fatal case where a child presented with severe diarrhea, followed by seizures, coma, and death. A sibling had died of the same symptoms two weeks earlier, and neither child showed any respiratory symptoms characteristic of influenza. Nevertheless, H5N1 virus was isolated from several specimens isolated from the second patient. Therefore, we are likely missing even many symptomatic infections--because they are presenting with the wrong symptoms. It is worrisome to think that the virus may be adapting to humans during infections like these. Additionally, avian influenza has been shown to cause asyptomatic infections. A report in the October Journal of Infectious Diseases showed that in a group of workers exposed to avian H7N1 and H7N3 influenza viruses during the Italian outbreak of 1999-2003, 7 of 183 people tested were seropositive to at least one of the viruses, suggesting they had been infected during the epidemic. None reported a history of influenza-like illness following exposure to the avian viruses. It's not known what the extent of sub-clinical infection with these viruses may be, either in Asia, or in the United States or Europe.

An influenza pandemic is a kind of viral perfect storm. Three requirements must be met: 1) the human population must lack antibodies to the virus; 2) the virus must make humans ill; 3) the virus must be efficiently transmitted between humans. H5N1 certainly meets the first two requirements; the scientific community is holding its collective breath waiting for the third condition to be met. In tomorrow's article, I'll discuss more about surveillance efforts and models used to predict or control an outbreak, should one occur.

Robert Webster is currently in his mid-seventies, and much of that life has been spent battling and studying influenza. With all he's seen, it's significant that he's made the following comments regarding H5N1:

This is the worst flu virus I have ever seen or worked with or read about. We have to prepare as if we were going to war--and the public needs to understand that clearly. This virus is playing its role as a natural bioterrorist. The politicians are going to say Chicken Little is at it again. And, if I'm wrong, then thank God. But if it does happen, and I fully expect that it will, there will be no place for any of us to hide. Not in the United States or in Europe or in a bunker somewhere. The virus is a very promiscuous and efficient killer.

This is the "war on terrorism" that we really should be sinking money into. Even if H5N1 isn't the next pandemic strain, or if it is but doesn't come anywhere near the level of the 1918 outbreak, it has served to point out holes in our preparation procedures for a disaster of that magnitude. More on that tomorrow.

Other posts in the series:
Day 1: History of Pandemic Influenza.

More resources on pandemic influenza: