Preparing for Avian Flu

The rapid growth of a particularly nasty strain of avian influenza has world health officials very worried. Little can be done to halt the disease's spread, and unless nations do more to prepare for a possible pandemic, the death toll from this outbreak could be very large.

Already this avian influenza is endemic in poultry and wild birds in 10 Asian countries, and it has killed at least 49 people. Most infectious disease experts believe that a worldwide influenza pandemic is almost inevitable within the next few years. The United Nations World Health Organization has urged all member states to make pandemic preparedness plans and begin implementing them. An expert group convened by WHO last November concluded, "The unpredictability of influenza viruses and the speed with which transmissibility can improve means that the time for preparedness planning is now."

Last August, the U.S. government unveiled a draft pandemic preparedness plan. But some public health experts consider it highly flawed. They claim that it puts way too much emphasis on vaccines, which take at least six months to produce and can be stockpiled for only about 18 months, and too little emphasis on antiviral medicines that can be stored for five years. They say that once a pandemic starts, probably in Asia, it is likely to reach the U.S. in less than three months, long before large supplies of vaccine are available.

The U.S. plan lays out no concrete strategies for allocating limited supplies of vaccines and antiviral medicines to workers--such as health care, transportation, and post office personnel--who are needed to keep essential services operating. Some say the plan should be more like that of the U.K., which contains very specific strategies for distributing medicines and maintaining services during a pandemic. "If you have antiviral medicines or even a vaccine, you won't have enough. You are not going to be able to distribute this to everybody. So you need to have a plan," says David M. Ozonoff, professor of environmental health at Boston University School of Public Health.

THE VIRUSES involved in all pandemics originate in birds because birds are the natural reservoir for flu viruses. Since the early 1500s, the world has experienced 22 pandemics, or about four every 100 years, says Michael T. Osterholm, director of the Center for Infectious Disease Research & Policy at the University of Minnesota, Twin Cities. During the 20th century, pandemics occurred in 1918, 1957, and 1968. The 1918 Spanish flu is estimated to have killed 50 million to 100 million people worldwide and more than 500,000 in the U.S., exceeding the number of American combat deaths during the century. Globally, the 1957 Asian flu caused 1 million to 2 million fatalities, and the 1968 Hong Kong flu was responsible for 700,000 deaths.

Unlike ordinary cyclical influenzas, which tend to spare young adults, the 1918 pandemic killed more people in the 18-to-35 age group than in any other age group, Osterholm says. Part of the explanation is that this group had no immunity to the virus, in contrast to older people who may have acquired some immunity from an 1880 pandemic. Another reason, Osterholm says, is that a gene in the 1918 virus caused a "cytokine storm"--a flood of molecular messengers triggering inflammation in the lungs. Many T cells known as cytokines rush to the lungs, inflaming the tissue and crowding and sometimes obstructing air passages. Young adults' robust immune systems actually make them more susceptible than older people to cytokine storms, he explains. That is why during the 1918 pandemic, young people often died within two or three days of contracting the disease. Present-day avian flu victims have also experienced cytokine storms, he says, and nearly all the victims have been under the age of 35.

IN GENERAL, three conditions are necessary for a pandemic. A new flu virus--one that has never infected humans--must emerge from animal reservoirs. The virus, unlike most that infect animals, must possess the ability to make people extremely sick. And infected individuals must be able to transmit the microbe to others with casual contact. So far, two of these conditions have been met for the avian flu.

The virus now causing avian flu--designated H5N1--first showed up in Hong Kong in chickens and humans in 1997. Six people died of the disease. Millions of chickens were culled, and it appeared that the virus had been eliminated.

But H5N1 reemerged in South Korea in December 2003. There, it migrated to five provinces, leading to the deaths of 1.3 million chickens and ducks from infection or culling. No humans were known to be infected.

By January 2004, the disease had spread to 23 of 64 provinces in Vietnam, infecting chickens and several humans, six of whom died. Outbreaks were also reported in poultry and humans in Thailand, and poultry deaths from H5N1 occurred in Cambodia, Laos, and China.

In February, Indonesia reported an outbreak of H5N1. According to WHO, this was the first time avian influenza has infected livestock in that country. During February, the virus continued its spread through Thailand, Vietnam, China, Laos, and Korea, moving to 52 of Vietnam's 64 provinces, 40 of Thailand's 76 provinces, and much of China. By the end of February, the number of confirmed human cases of bird flu in Thailand and Vietnam was a combined 25, of which 19 were fatal.

From February 2004 on, the disease continued its trek across Asia, infecting not only domestic chickens, ducks, and geese, but also wild ducks and geese and a host of other wild bird species.

As C&EN went to press, H5N1 had caused a total of 74 confirmed human cases, of which 49 were fatal. A handful of the cases seem to have resulted from human-to-human transmission. There was one clearly documented case in Thailand a few months ago, in which a mother was infected, not from a chicken, but from her child who had been infected by a chicken, says Anthony S. Fauci, director of the National Institute of Allergy & Infectious Diseases (NIAID) (N. Engl. J. Med. 2005, 352, 333). Human-to-human transmission of H5N1 is still rare, but bird-to-human transmission seems to be increasing, he says.

WHEN H5N1 first showed up in South Korea, hopes were high that the disease could be halted by culling all the chickens on affected farms, and millions of animals on commercial farms were killed. But after the disease spread to wild bird populations, and to areas where the economy is dominated by subsistence farmers whose primary source of livelihood is small flocks of chickens or ducks running freely in backyards, culling became impractical as a way to eliminate H5N1. Culling would destroy the mainstay of much of, for instance, Thailand's population, and almost as soon as families brought in new flocks of uninfected chickens, these would be infected by wild birds. In large rural areas of Thailand, nearly every family has a flock of chickens and ducks. Infection among bird flocks in Asia is virtually out of control, Fauci says.

There is a vaccine for chickens against H5N1. But some people are concerned that if the vaccine is used extensively, the virus will evolve, making the vaccine useless. Or they worry that chickens will become infected yet show no symptoms. Under those circumstances, people could be exposed to the virus and not know it, Fauci says. But the problems resulting from not vaccinating chickens may be greater than the risks of vaccination, he explains.

The H5N1 virus is highly unusual in several respects. Most flu viruses affect a single species, but H5N1 has infected a wide range of animals: chickens, ducks, swans, storks, turtledoves, eagles, tigers, leopards, domestic cats, pigs, mice, and humans. No other influenza virus has spread so quickly over such a wide geographical area, Fauci says. And so far, H5N1 has shown extremely unusual virulence, killing about 67% of the humans infected. "It is totally unprecedented to have bird flu in so many countries and so many farms and in so many different species," says Klaus Stöhr, head of WHO's global influenza program.

These apparently unique traits worry experts. They say that each new human case of avian flu provides an additional opportunity for the H5N1 virus to reassort--exchange genetic material--with a human flu virus and to acquire traits that could allow it to spread easily among humans. They fear that if the virus shifts so it can be transmitted among humans with casual contact and retains even a small fraction of its current virulence, it will pose a grave threat. In most years, ordinary, cyclical flu kills fewer than 1% of those infected.

The flu pandemic in 1918 killed 1 to 3% of its victims in the U.S. The number of deaths was extremely high because people had virtually no immunity to the virus, and it infected about one-quarter of the global population. "The situation in Asia is very concerning. We have an avian influenza virus which is very widespread but which has not yet, fortunately, transmitted easily, rapidly, and for a long period of time between humans," Stöhr said last September.

"It would be folly to assume a pandemic is not going to occur," Fauci says. Experts warn that if a bird flu pandemic strikes the U.S., the number of fatalities could easily be several times the 36,000 seen in a regular influenza year.

LESSONS FOR TODAY
Swine Flu Debacle Could Affect Pandemic Planning

Just as the Vietnam War experience engenders caution in U.S. war strategists, the debacle over swine flu vaccinations nearly 30 years ago is on the minds of those planning for potential influenza pandemics.
The swine flu disaster has lessons for today, however.
In January 1976, several people in Fort Dix, N.J., came down with a swine influenza virus strain possessing antigen subtypes that had not circulated for 50 years. At the time, many scientists believed that introduction of a new virus subtype that was capable of being transmitted from person to person—even if there was only a small cluster of cases—nearly always resulted in a pandemic.
An advisory committee met and decided that human-to-human transmission of the swine flu virus had occurred, but there was no way to tell whether it would cause a pandemic.
Nevertheless, on March 24, representatives from the Centers for Disease Control & Prevention (CDC), the Food & Drug Administration, and the National Institutes of Health recommended mass immunization of the U.S. population.
Swine flu immunizations began on Oct. 1, and in the following 10 weeks, more than 40 million people were vaccinated, compared with only 10 million during the previous flu season.
But the plan did not go well. In November, several cases of Guillain-Barré syndrome were reported in Minnesota. Over the next month or so, a total of 532 cases and 32 deaths from the syndrome were reported throughout the U.S. CDC investigated and decided that vaccination with the swine flu vaccine increased the risk of Guillain-Barré. On Dec. 16, the vaccination program was terminated. After the initial cases of swine flu at Fort Dix, no additional victims were found.
Officials have since decided that several mistakes were made during this episode.
Decisions made early in the process were not modified by new findings—specifically that no new swine flu cases were detected after the Fort Dix outbreak. Severe unexpected adverse reactions—Guillain-Barré syndrome—occurred in some people who were vaccinated, but no system had been set up ahead of time to deal with adverse reactions. Another problem was that implementation of the vaccine program by the states varied. If there had been an epidemic, some states would have been unprepared. And if there had been a severe epidemic, vaccine production would not have been great enough to protect the entire population.

ALTHOUGH AMONG known H5N1 human cases, the fatality rate is about 67%, it may actually turn out to be much lower than that when more careful surveillance is done. Recently, the virus has been isolated from people who have shown few or no symptoms.

There are undoubtedly many cases of H5N1 that have gone unreported, says Dick Thompson, media coordinator for communicable diseases at WHO. "First of all, there is probably a wide range of manifestation of illness. Some people get so sick that they go to a hospital. These are the people we know about," he explains. "But during epidemiological investigations, we have discovered others with milder symptoms who do not go to a hospital. That is the situation in Vietnam, and it is likely that this same sort of thing is occurring in Cambodia or elsewhere where the surveillance system is not very good."

In many countries, the primary strategy for defeating a potential pandemic is to stockpile the drug oseltamivir, trade named Tamiflu, made by Roche Pharmaceuticals. Tamiflu is a neuraminidase inhibitor that can prevent many types of influenza if taken right before exposure, and can reduce the severity of symptoms if taken within 48 hours of infection, Boston University's Ozonoff says. Altogether, 15 countries have ordered Tamiflu, and Roche is negotiating with about a dozen more. According to Roche, in vitro tests show that Tamiflu is active against about 3,000 types of influenza viruses.

Some of these countries are putting a lot of confidence in this drug. Norway is stockpiling enough Tamiflu to treat one-third of its population, and the U.K., France, and New Zealand have ordered sufficient quantities for roughly one-fourth of their populations. Australia has ordered enough for about one-fifth of its citizens. In contrast, the U.S. is buying 2.3 million treatments, which could medicate less than 1% of Americans.

Roche will not reveal its total production capacity for Tamiflu or the price being paid for the stockpiled drug. "There is one single world price that all countries are paying for the Tamiflu being stockpiled," and this price is less than the price for Tamiflu that will end up in pharmacies, Roche spokesperson Martina Rupp says.

"It will take several years for Roche to fulfill the orders placed so far," because they greatly exceed the company's yearly production capacity," University of Minnesota's Osterholm says.

Thailand is considering manufacturing its own generic supply of Tamiflu. "This would be difficult," Rupp says. "The manufacturing process takes up to 12 months and is very complex, involving one potentially explosive step."

Ozonoff is highly critical of the U.S. Tamiflu policy. "We should have stockpiles of antivirals," he says. "The U.S. has ordered a pitiful amount from the manufacturer. It ought to be investigating ways it can make more--manufacture Tamiflu if Roche doesn't have the capacity to do it."

The U.S. pandemic flu preparedness plan relies almost entirely on vaccines. Eight thousand doses of a flu vaccine against H5N1 have been manufactured, and clinical trials of this vaccine are beginning, Fauci says. The purpose of the trials is to test for safety and dosing--to see whether one or two shots are required for an appropriate immune response. The trials will initially be done on about 450 young adults. Then, if that goes well, trials will begin on a population of seniors and eventually on a group of children, he says.

In parallel with the trials, Fauci says, Sanofi Pasteur is manufacturing 2 million doses of the H5N1 vaccine for the U.S. national stockpile. The reason for scaling up immediately to 2 million doses is so the operation can be easily expanded to produce 20 million to 40 million doses if that is necessary, he explains.

One holdup is that Sanofi Pasteur must finish manufacturing the ordinary trivalent vaccine for the 2005-06 flu season. It will finish that process in late July, and then it can start producing tens of millions of doses of avian flu vaccine. The entire manufacturing cycle for any flu vaccine takes at least six months.

Osterholm is not so sure that Tamiflu would work in a pandemic. "We assume it will have some efficacy," he says. "Tamiflu works by reducing and in some cases stopping replication of the virus. But on the cytokine storm response we are seeing in patients in Southeast Asia, we don't know how Tamiflu would work."

THE U.S. PANDEMIC preparedness plan is far too general, says Walter E. Stamm, president of the Infectious Diseases Society of America (IDSA) and professor of medicine at the University of Washington, Seattle. "We need to develop a plan that has considerable detail in how we actually coordinate pandemic responses," he says. "The Department of Health & Human Services should develop a detailed plan to coordinate pandemic response at all levels, from local to state to national, including links between federal authorities and clinicians throughout the country."

Another of Stamm's recommendations is to stockpile enough antiviral drug to medicate at least 50% of the U.S. population--about 150 million courses of treatment. He also wants the government to create incentives for pharmaceutical companies to get back into the vaccine business and increase capacity for vaccine production. "There needs to be a guaranteed market. The government needs to commit to buy a certain number of doses no matter what happens," he says.

"One of the things we would have to deal with in a pandemic is that we live in a just-in-time delivery economy," Stamm says. "Everything from mechanical ventilators to masks to caskets: All these things would be in short supply if we have a pandemic. We have to plan around that."

IDSA has been meeting with congressional leaders, trying to convince them of the need for immediate action.

Ozonoff has created a pandemic to-do list that is similar in some respects to Stamm's recommendations. Because there aren't enough hospital beds for even a moderately bad epidemic, "we need to set up alternatives, such as hotel and motel rooms to use in a pandemic," he says. In a pandemic, the nursing shortage would be particularly acute, especially if many nurses become sick with influenza, he warns. Some alternatives include organizing retired nurses and training volunteers, he says.

FURTHERMORE, inventories of respirators need to be acquired, Ozonoff says. "Currently, there are not enough respirators to handle a huge increase in adult cases, and certainly not enough pediatric respirators to handle a disease with a proclivity for the young."

In early-17th-century England, authorities sometimes quarantined plague-struck families in their houses and tenements by nailing the doors and windows shut. Fears that disease would spread faster if victims ventured outside condemned those inside to an almost certain death. Medical science has advanced spectacularly since then; yet in some regions of the world, the technical capacity and equipment needed for preventing the spread of infectious disease have not advanced much from what they were in Shakespeare's day.

For example, in March, WHO had to fly dry ice from Bangkok to Cambodia to keep avian flu virus specimens cool during shipment to the Pasteur Institute in Paris. Lab technicians in Southeast Asia often work under conditions that make it extremely difficult to obtain accurate results. Many veterinarians in Vietnam lack telephones or fax machines to communicate with the rest of the world and report disease outbreaks.

Unlike during the 1918 pandemic, scientists today have the knowledge needed to prepare vaccines and are aware that a pandemic may be brewing in Asia. Yet most of the wealth and facilities to produce vaccines are in developed countries, and the handful of firms that make them can manufacture enough for only a fraction of the world's population. If a pandemic were to occur, 2.5 billion to 3 billion doses of vaccine would be needed, but current production facilities could make at most a total of 300 million doses, Osterholm explains. In the past, vaccines have never been available early enough and in sufficient quantities to have an impact on morbidity and mortality during a pandemic.

For the first time when the world has faced a pandemic, an antiviral medicine--Roche's Tamiflu--is available that may be effective in preventing infection and ameliorating symptoms. Yet only a relatively few countries can afford to stockpile Tamiflu, and they may not be purchasing nearly enough to protect their citizens. Even if they did order enough, Roche would not have the capacity to fill the orders.

Every day, individuals from countries where surveillance for bird flu cases is poor, or nonexistent, fly to places where the disease has not yet arrived. The combination of promising treatments, highly unequal technical skills and equipment, highly unequal wealth, and thousands of people traveling from Asia to Western countries each day is unprecedented on the eve of a potential pandemic. Poor planning, an absence of political will, and sheer lack of wealth and technical expertise may be the nails that condemn millions of people to unrelieved symptoms and unnecessary death.