The flu hunters: special report

They scour the earth for clues that could prevent millions of deaths. But can the world's leading flu scientists stop the next lethal pandemic?

Early this September, a death announcement appeared on Disease Outbreak News, a World Health Organisation website that typically contains the first official report of unusual flu cases. A young man had died in Thailand the previous day, of avian flu. "The case," the posting read, "was an 18-year-old male who'd been exposed to sick chickens."

Early this September, a death announcement appeared on Disease Outbreak News, a World Health Organisation website that typically contains the first official report of unusual flu cases. A young man had died in Thailand the previous day, of avian flu. "The case," the posting read, "was an 18-year-old male who'd been exposed to sick chickens."

In a follow-up posting that day, on ProMED-mail, the infectious-disease world's invaluable weblog, things became more intriguing. The posting, originating from the Thai Department of Disease Control, said the young man had had "very close contact to... fighting cocks by carrying and helping to clear up the mucous secretion from the throat of the cock during the fighting game by using his mouth".

That is, the young man had cleared the cock's airways by sucking his rooster's beak and swallowing the spit and mucus. "That was a risk factor for avian flu we hadn't really considered before," Dr Tim Uyeki told me dryly.

Uyeki is a leading epidemiologist at the US Centres for Disease Control and Prevention (CDC). Few people in the world know as much about how influenza can be expected to behave as Uyeki and his boss, Dr Keiji Fukuda. When Fukuda admits to being as concerned "as I have ever been", people who know him start really worrying.

Fukuda's concern is well founded. In the past year, he has watched from his office in Atlanta as events overseas have seemed poised to spiral out of control. Between January and the end of October, 32 people have died from avian influenza in Vietnam and Thailand. Tens of millions of chickens have died and millions more have been slaughtered. More nations have admitted to outbreaks among birds in more provinces than would have been conceivable even 18 months ago. All this, Fukuda says, "certainly increases the possibility" of a much larger outbreak of avian flu among people.

As the point man for day-to-day updates on avian flu in Asia, Uyeki reads and files the postings and checks in with colleagues on the ground overseas. This is a huge benefit of his having done fieldwork in several countries. "I know the scientists in many parts of Asia," he says, "and they know me."

Epidemiology is a beat reporter's science. Those who are good at it collect bits of information and unofficial rumours. They cultivate confidential sources. These flu hunters are expected to forestall any big outbreak, but the real fear for Fukuda and Uyeki is that they might not know one is starting until it is too advanced to stop.

Events in Asia are escalating almost daily, with more human deaths, more species being infected and more questions arising about how the virus there behaves. At the same time, some Asian governments provide little information. This reticence greatly increases the difficulty of the epidemiologists' work.

Fukuda and Uyeki proceed from a position of pessimism. They live with the spectre of the 1918 flu pandemic, which killed between 20 million and 50 million people worldwide. They live with the assumption that another pandemic is inevitable.

The announcement about the Thai teenager's death, made by the country's Department of Disease Control, said also that the national and local health authorities in the young man's province would study the case. Village chickens would be slaughtered. The victim's neighbours and the doctors and nurses who had cared for him would be put under close observation in case he had passed on the infection.

Uyeki felt stymied, because he had never done fieldwork in that country, and because avian flu is such a potent issue there. Until this year, Thailand was the world's fourth-largest chicken exporter. Its industrial poultry producers employ hundreds of thousands of people. But after it reported on 23 January to the World Organisation for Animal Health (WOFAH) that avian influenza had been detected in a poultry flock, nations began to bar the import of its birds. There were calls for the prime minister and other officials to resign.

As the weeks went by this autumn, the Thai government did not release any further reports on the cockfighter's death. The previous winter, as Uyeki knew well, people in Vietnam and Thailand had fallen ill at about the same time as chickens had. Within the space of two months, at least 35 people had developed severe respiratory distress, and 23 of them had died.

But new human infections mysteriously sputtered and then ceased. The last death was in Vietnam in March. "There had not been any obvious sustained person-to-person transmission" in the outbreak, Uyeki says; a prerequisite for a pandemic.

Then, in June, Indonesia reported avian influenza in poultry to WOFAH. Vietnam and Thailand followed in July, then China, Cambodia and Malaysia. And the virus popped up again in humans; there were reports of infections in Vietnam, with the first death in early August. Two more followed in Vietnam - and then the cockfighting teenager in Thailand died on 8 September. Uyeki craved details: "Any time there's evidence of a developing outbreak, you want to know as much as possible."

On 27 September, the Thai authorities reported that a 26-year-old woman living in Bangkok's suburbs had, it seemed, caught avian flu from her 11-year-old daughter. The girl, who had been living in a village with her aunt, had helped to dispose of sick chickens, and then fallen ill and been hospitalised. Her mother had rushed to tend to her child. The girl died. The day after her funeral, the mother started to feel ill and, after severe illness, died. The girl's aunt got sick too, but survived. Both mother and aunt tested positive for avian influenza.

Uyeki avidly read the ProMED postings, and called and e-mailed colleagues. Soon, he had the exact dates of the onset of disease, hospitalisation and death. He learnt that the mother had spent about an hour in the village before going to the hospital. Had she picked up a chicken, touched a knife, walked through a pen?

"Possible person-to-person transmission grabs our attention," he says. When a virus can spread easily among people, it can move rapidly around the world, arriving anywhere within the space of one airplane ride. "It's something that needs to be thoroughly investigated," Uyeki says.

Fukuda joined the CDC in 1996. Until then, influenza had been a backwater in infectious disease. "Diseases such as HIV had the world's attention," Fukuda says. That was about to change.

A year later, he was completing a clinical rotation at a hospital in San Francisco. It was August 1997. He got a phone call saying that a three-year-old boy had tested positive for avian influenza in Hong Kong. It was "a highly unusual event," he recalls.

To appreciate what a staggering piece of news this was, you'll need to understand something about the characteristics of flu. Influenza viruses are divided into three types, A, B or C, depending on the virus structure. Humans can be infected with all three, although C-class flus are uncommon and B doesn't usually cause severe illness except in children.

Influenza A is the monster, in animals and people, usually causing the most virulent illness. Type A influenzas (unlike Bs or Cs) have multiple subtypes with cryptic names: H3N2, H1N1, H7N2. The "H" is for haemagglutinin, a spiky protein on the surface of the influenza molecule. In human flu viruses, the spikes of haemagglutinin connect, like sinister Lego blocks, with matching receptors on the outside of healthy respiratory-system cells. The virus then melds with the healthy cell and begins replicating. Neuraminidase, the "N" in the flu name and another protein, uncouples the virus from its host, tearing the cell membrane, allowing the progeny to escape and killing the cell. Loosed, they start repeating the process deeper and deeper into the respiratory tract.

The little boy in Hong Kong seven years ago had a type A influenza, H5N1, a bird flu. H5N1 wasn't new; it had been isolated in wild birds as long ago as the early 1960s. But this avian flu had never been known to infect a human being. An avian flu can occasionally jump from a bird to a person. Nature is fluky. The resulting infection is a nasty shock for the patient, but not much of a threat to the rest of us.

In order to cause a pandemic, an entirely new human influenza subtype must emerge, one most people would never have encountered and would not have immunity to. It must also be able to be transmitted efficiently from person to person. In recent history, three main types of influenza A have circulated freely in humans. (Many more exist in birds.) Each of these strains caused a pandemic. The catastrophic 1918 Spanish flu pandemic was caused by H1N1. The 1957 Asian flu pandemic, which killed millions worldwide, was brought about by H2N2. And the 1968 pandemic of Hong Kong flu introduced H3N2.

Science doesn't quite understand how a non-human flu virus adapts and becomes a human flu - or why more don't. But they do suspect it is a process of gene swapping. If an avian flu infects a person who, coincidentally, suffers from a human flu, the two bugs might exchange genetic material. The resulting virus could be essentially avian, but with genetic components of the human flu that allow it to be easily transmitted from person to person. This is how the Asian and Hong Kong flu pandemics began.

Shortly after hearing about the three-year-old boy, Fukuda took part in a conference call between the CDC and the Hong Kong health department. He was, he says, trying to keep his anxiety but undeniable excitement in check. "There was a surge of adrenalin. But at that point, we had one case - an extremely interesting case, but only one." He and several CDC colleagues soon left for Hong Kong.

For an epidemiologist, fieldwork is gruelling, grimy and at times dangerous. It is also, when everything clicks, the surest way to understand the progress of a disease and, in the best of circumstances, to stop it cold.

When Fukuda arrived in Hong Kong, he studied the dead boy's medical charts. It had been a hard death, which shook Fukuda, the father of two young children. The child had a breathing tube inserted and was in great pain. "As I went over his chart, the case became very un-abstract," he said. "It drove home for me how much suffering there might be if this bug took off."

Fukuda soon learnt that chickens had been dying in the New Territories region of northern Hong Kong. Routine testing of the poultry had turned up H5N1 infection.

Fukuda had his local colleagues ask the boy's parents if they had visited the affected region lately. It turned out that the boy had attended day care and the teachers had kept chicks as playthings for the children. Some of the birds had died. None now remained, so no testing of the animals was possible.

Although Fukuda surmised that the boy caught the disease from a chicken with H5, the epidemiologists never could establish with certainty how the boy had become ill. There were many unanswered questions: why didn't other children get sick? Had this boy played differently with a bird? At a different time? Fukuda returned home. "There was a sense, I guess, of incompletion," he says, "of dissatisfaction and some concern. It simply wasn't clear at that time whether this had been a unique, transitory event."

Then, three months later, an e-mail arrived at the CDC from Hong Kong. There was another H5 case. The authorities asked Fukuda and his colleagues to return.

By now, it was early December. By the time Fukuda arrived, a second human case of avian flu had been confirmed in this new outbreak - then a third and a fourth. Fukuda and his team of seven from the CDC set up a war room. Each day, they tallied new cases. The total kept growing. "None of us were sleeping much," Fukuda recalls. "The adrenalin was really flowing at this point. A pandemic was suddenly not a misty historical possibility. It seemed very current."

The investigators mapped out the homes and haunts of the sick. There was little overlap among the flu victims. The contagion didn't seem to be passing from one person to another. The team checked with its agricultural counterparts. Vets had found evidence of a recent outbreak of H5N1 in dying chickens in one part of Hong Kong, but few, if any, of the human patients had visited that region. However, most of them had been to a local live poultry market the week before falling ill.

On 28 December, Fukuda and health department chiefs were called to an emergency meeting. Their veterinary counterparts had been working on a parallel investigation, taking samples of bird faeces and blood. The findings chilled those assembled; H5N1 appeared to have spread throughout the nation's live poultry markets.

"You have to understand this news within the context of the human disease situation," Fukuda says. Small outbreaks among poultry in May had preceded the first case. Now the infection was widespread. "The solution was obvious but not easy," Fukuda says - Hong Kong should slaughter every bird within its borders.

There was hesitancy. If chicken markets were not the direct cause of this outbreak, killing poultry would be an enormous economic waste. If, however, Hong Kong opted not to kill the birds and the disease spread... "When looked at that way, the conclusion became clearer," Fukuda says.

Government ministers were called. They didn't balk; within hours, preparations started. Poultry markets were closed and disinfected. Culling began the next day and was completed in three days. Overnight, new human infections ceased. "I remember it as the most satisfying investigation of my life," Fukuda says.

On 29 December last year, Uyeki was at his desk when he received an e-mail that made him blink. It was a private, informal request for advice from a virologist at the National Institute of Hygiene and Epidemiology in Hanoi, Vietnam. "We have some children [sic] patients with respirator symtomes [sic]," she wrote. "We need to know what's causative of it." Several children from all over Vietnam had become so ill that they had been brought to a paediatric hospital in Hanoi.

Uyeki has a particular interest in the effect of influenza on the young. It is one of his specialisms. In the months just past, a number of children in the US had died after getting the flu. He thought that Vietnam might be dealing with the same virulent human flu and suggested that the virologist test the youngsters for human influenza.

He might have urged her to test for bird flu had he had confirmation of any substantial numbers of chicken deaths in the region. There had been rumours circulating among influenza scientists that chickens were sick in Vietnam, Thailand and elsewhere. But no Asian country, except South Korea, had reported poultry deaths from influenza to WOFAH. Reporting to this world organisation is strictly voluntary. It is also inevitably self-punishing, usually triggering export restrictions.

The silence from Asia didn't necessarily mean that people were covering up chicken flu there. Most Asian nations don't have sophisticated veterinary surveillance systems for detecting and reporting sickness in animals. Perhaps, too, "people don't want to know," Fukuda says, "because of the economic and political consequences."

That night, Uyeki received another e-mail from the Vietnamese virologist: "Dear Tim... these childrens... have fever, cough, difficult breath... some of them have diahrea [sic] after few days of Honset and died quickly." This was the first he had heard of deaths. Uyeki pressed her again to test further for human flu.

On 8 January, Vietnam reported for the first time that it had cases of avian flu in chickens, some of which were thought to have occurred as early as 27 December. Uyeki shipped the virologist specialised tests to check for a wide range of influenza A subtypes, including H5. A few days later, the children's illness was confirmed. They had H5N1, avian flu. By then, 11 children and one adult had died.

On 12 January, Japan reported to WOFAH that it had found H5N1 in one flock of hens. Eleven days later, Thailand reported that nearly 9,000 chickens on one farm were infected with H5N1. The following week, it reported that almost 10 million chickens were infected. That same month, Cambodia and Laos reported poultry outbreaks. China and Indonesia soon followed suit. Throughout that time, human cases continued to stream into hospitals in Vietnam, and then in Thailand. Vietnam asked the WHO to help investigate. On 18 January, Fukuda and Uyeki and four CDC colleagues left for Hanoi to join an international team which, over the next six weeks, grew to include scientists from Britain, France, Japan, Sweden, Holland, the Philippines, Switzerland and Vietnam.

One of the first concerns of epidemiology is why a particular person gets sick and not another. The inquiry is tougher when the flu victim is young or dead. In Vietnam, they often were both.

The ramifications of their investigation were also a stumbling block. Vietnam isn't a major poultry exporter. Chickens tend to be for local consumption. If avian flu is confirmed in a village, all the local flocks are slaughtered and burnt. Few villagers were eager to talk about sickness in their birds.

"It's easy when you're on the outside to say these people should do this or that," says Fukuda, who advised Uyeki during the investigation. "It's different when you're actually in Asia and realise that resources are so limited." But they pressed on. By the end of their stay, they had uncovered poultry contact for most of the sick people in the region. Then, in March, the outbreak simply stopped. It would inexplicably reappear in humans this summer and autumn.

"It's troubling to me that we still don't really know much more about this virus than we did in 1997," Fukuda says about the current state of avian flu research. His Hong Kong investigation proved that close contact with sick poultry was a significant risk factor for H5N1 infection. But there is still a long list of vital questions about the H5N1 virus.

Why haven't poultry workers become ill, for one? Hundreds of thousands of people labour in Asia's live-chicken markets or have helped in the chicken-culling process. "But hundreds of thousands of people have not died," Uyeki says. "Thousands haven't died. Thousands haven't even been hospitalised. Most of us in influenza think that there have been more cases and more deaths from H5N1 than we've heard about. But there haven't been huge numbers of undiscovered deaths."

"Where are the studies being done to learn more about transmission?" Fukuda asks wearily.

A 1997 study of blood samples taken from poultry workers in Hong Kong found that 10 per cent had developed antibodies to H5N1. In 2001, Vietnamese scientists, with help from Uyeki, took blood from 200 poultry workers. The data haven't been published or updated.

"We also don't know how the bird virus managed to infect so much of Asia so quickly," he says. Not long ago, H5N1 was confined to Hong Kong. But this year, it has spread through chicken flocks so far afield and so pervasively that the possibility of eradicating it any time soon seems slight.

There are other daunting questions. Is today's H5N1 changing? Influenza viruses mutate constantly; they are unpredictable. But H5N1 seems to be altering in disturbing ways, Fukuda says.

The flu strain has also infected and killed domestic cats and pigs. It swept through a zoo in eastern Thailand, killing more than 20 tigers. It has been shown to be transmittable from cat to cat, which no avian flu has been before. This movement into other species concerns flu hunters for many reasons, but principally because it suggests that H5N1 is becoming highly capable of infecting mammals. Also, some of the infected animals, such as swine, can catch both human and avian flus simultaneously. So the gene shuffling required to make the avian flu better adapted to humans could take place within a pig.

Fukuda and his colleagues also worry that, as H5N1 mutates, it will be difficult to produce an effective vaccine. In May, the US National Institutes of Health contracted with American companies to make an H5N1 human vaccine using the avian flu strain that killed people in Vietnam last winter. The vaccine is scheduled to go into trials early next year. But if the circulating H5N1 strain changes significantly, the vaccine could be rendered less effective.

In October, scientists in the CDC's flu labs genetically sequenced the virus that killed the 18-year-old Thai cockfighter and found that it remains structurally similar to the strain used in the vaccine. Fukuda says he feels reassured, at least for the moment.

But it's the worst-case scenarios that haunt Fukuda, Uyeki and their colleagues. "There's so much reluctance in Asia to look for avian flu," Fukuda says, so the particular moment when an epidemic passes into a pandemic may go unnoticed.

The first sustained human-to-human transmissions might start insidiously, perhaps within one family whose infections aren't immediately identified as H5N1. It could spread through their village, to the nearest city and then around the world, leaving science no chance to catch up.

Some scientists outside the CDC are less concerned about H5N1 mutating into a true human flu. If it could, it would have done, they argue. There may be something in its genetic structure, they say, that makes the move impossible. Fukuda shrugs. "That would be nice, wouldn't it?" he says. "We have no way of knowing."