Feature: Avian flu
9 October 2005. By Giles Newton, science editor at the Wellcome Trust.
We may think we have flu when feeling under the weather, but the real thing is much worse. We are likely to be completely flattened for at least a week. But unless we are already weakened or vulnerable, we will recover. However, for the old and frail, flu is highly dangerous. In the UK, it kills some 12 000 people every year.
Striking though this number is, it pales against the impact of global flu pandemics. Three times in the 20th century, pandemic influenza swept across the globe. These pandemics not only killed the very old, the very young and the infirm, but also cut a swath through fit and healthy individuals.
In recent history, flu pandemics have arisen every ten to fifty years - so where will the next one come from?
One leading candidate is avian flu H5N1, and it has the experts extremely worried. For a bird flu to 'jump the species barrier' and infect humans is rare, but H5N1 has done so to shocking effect: it has killed around half of those infected. The great fear is that H5N1 will become established in people and be transported around the world.
Avian flu H5N1
Influenza viruses tend to go for monogamous relationships: human viruses stick to humans and avian viruses to birds. The bird viruses are normally found in wild ducks and other migrating waterfowl - the natural 'reservoir' - which, along with domestic ducks, usually show no obvious effects. But H5N1 is unusually nasty to birds such as domestic chickens: a flock of 10 000 can be annihilated within hours.
H5N1 came to the world's attention in 1997, during an outbreak among Hong Kong's poultry. Eighteen people were also infected, of whom six died. Hong Kong destroyed its entire poultry population of 1.5 million birds in just three days and a potential pandemic was averted. Even so, H5N1 re-emerged in late 2003 and early 2004, in an unprecedented epidemic sweeping through poultry in most Asian countries. Tens of millions of birds were killed by the virus or culled. In January 2004, the first cases of infected people were seen in Vietnam; by 17 June 2005, at least 104 people had been infected in Vietnam, Thailand and Cambodia, and 54 people had died.
H5N1 causes a massive destruction of lung tissue: an infected person can seem to have mild pneumonia in the morning, yet be dead by mid-afternoon. Almost all of the people affected have been in close contact with chickens or ducks. Most families in South-east Asia own birds, which tend to occupy the same living space as people, creating conditions in which people are constantly exposed to animal pathogens. Birds are also so economically important that people will do almost anything to keep their stocks of chickens and ducks.
Nevertheless, H5N1 still seems relatively inefficient at infecting humans. The real dangers would come if the virus adapted and began to be transmitted from person to person. So far there has been one confirmed case report of human-to-human transmission of H5N1: a Thai woman who contracted it from her daughter. However, there have been other suspected cases where limited human-to-human transmission was the probable route of transmission.
Ironically, the virus will also be a bigger threat if it becomes less deadly: with more people alive but infected, the virus has more opportunity to spread. Spanish flu, for example, killed only 2 per cent of those it infected, but infected millions.
These changes may come about in two ways. Like any RNA virus, influenza makes mistakes while replicating its genome, creating considerable variability in its genes and proteins. The recent Asian H5N1 strains are different from the 1997 Hong Kong ones, and continue to gain the ability to infect new species. Migratory wildfowl are no longer immune to its effects, pigs have been found with H5N1 for the first time, and tigers at a zoo near Bangkok died after being fed infected chicken carcasses.
Alternatively, H5N1 may swap genes with human influenza virus - as happened in 1957 and 1968. This is most likely to occur in pigs, genetic 'melting pots' that can be infected with both avian and human viruses.
Halting H5N1
So what can we do about H5N1? Stopping it at source is vital, and millions of poultry have been culled in Asia since late 2003. Thailand, for example, has already cut H5N1 in ducks to undetectable levels. But culling is unlikely to be a permanent answer: millions of poultry remain in Asia.
What about vaccines? The main problem here is the virus's variability: vaccines only work against certain virus strains. The strains to be included in the standard, 'seasonal' vaccine are chosen each year in February. There is no guarantee, however, that the chosen strains will be those that eventually predominate.
Bizarrely, the vaccines are still produced in chicken eggs. Scaling up vaccine production will therefore not be straightforward. Although about 300 million doses are made each year, billions of doses might be required if a pandemic arises.
Prototype vaccines that offer protection against H5N1 have been produced. While some governments have ordered them, others have held back - as these vaccines might not counter the strain causing the pandemic.
Do flu drugs give us any more hope? Oseltamivir (Tamiflu) and zanamivir (Relenza) are neuraminidase inhibitors that can reduce the release of viruses from infected cells; they work well if taken within two days of infection. Two older drugs, amantadine and rimantadine, also work well but can have unpleasant side-effects and resistance can arise rapidly (the 2004 strain of H5N1 is resistant to amantadine). Resistance is likely to be a problem with oseltamivir: a single nucleotide change in the virus genome can render it resistant. And with only one class of drug available, combination therapy is impossible.
If that were not bad enough, the course of a flu epidemic creates yet more problems. Once begun, an epidemic peaks in a matter of weeks, before dying away as the reservoir of susceptible victims disappears. We would have perhaps a fortnight to do anything to limit its impact.
Even more soberingly, Spanish flu killed about 50 million people in 1918-19, but today's global population is much larger, with huge urban centres and rapid air travel. Even with modern healthcare, experts estimate that anywhere between 2 million and 50 million people could die in a pandemic.
The WHO has been urging countries to develop or update their influenza pandemic preparedness plans. The UK Government announced its plan in March 2005, estimating that a flu pandemic could affect about a quarter of the population and cause more than 50 000 deaths. Then in July 2005, it said it would acquire two million doses of flu vaccine along with 14.6 million doses of oseltamivir (enough for a quarter of the population, as recommended by the WHO), and could introduce quarantine measures and travel restrictions, as well as cancelling concerts and football matches, to stop the virus spreading. Restrictions on movement and other interventionsin 1918, however, probably had no effect on the scale of the pandemic.
