Flu has laid Britain low this winter. Would an epidemic have been prevented had government advisers given support to a possible wonder drug? Julia Hinde reports.

If I got flu, I would want Relenza," says the Australian architect of the much-hyped anti-flu drug that government advisers decided not to recommend to British doctors for prescription, but that is flying off chemist's shelves in the United States. As thousands across Britain sweat and shiver with flu, Peter Colman, the physicist whose research provided the cornerstone for a new approach to beating the influenza virus, is confident the drug works.

"It is a hard thing to measure," says the head of Melbourne's Biomedical Research Institute, a joint government/industry research centre. "It is tough in a disease that is self-limiting. Most people get better from flu. I believe the drug's benefits have not been fully described."

The drug Colman helped develop hit the headlines late last year when the UK's new National Institute for Clinical Excellence, which advises the government on the clinical benefit and cost-effectiveness of new drugs, recommended that Relenza should not be prescribed by the National Health Service this year.

ÖйúAƬ

Ministers' decision to accept the recommendation was a blow to GlaxoWellcome, which had spent ten years developing the drug and hoped that it would be prescribed in huge numbers this winter. AstraZeneca and SmithKline Beecham also condemned the decision.

In its first judgement on a new drug, the Nice committee - comprising numerous top academic medics - concluded that though trials showed Relenza reduced the length of flu by 24 hours, there was insufficient information (because of the small number of patients) to judge how the frequency of serious secondary complications in high-risk patients might be reduced. With this in mind, Nice decided Relenza - likely to cost Pounds 24 for a five-day treatment - should be subject to further trials.

ÖйúAƬ

While recent talk of Relenza has centred on business interests and pharmaceutical company profits, Colman's search for a drug to fight flu started from much more humble beginnings more than two decades ago.

A physicist with no medical training, Colman took an early interest in the potential of physics to solve biological problems. He was intrigued by antibodies - molecules produced naturally in the body to fight infection. Colman wanted to understand how different a new virus needed to be before existing antibodies, produced in the body to fight a former infection, no longer recognised and attacked the new intruder.

"Flu is a great system for studying that," explains Colman, "because flu is always changing." That characteristic makes the Australian approach potentially so powerful, while also making flu itself so deadly.

Essentially genes wrapped in proteins, a type of flu virus has the ability to mutate regularly so that animals, which may have built up resistance to one strain of the virus, are unable to recognise a mutated form and therefore can become infected again.

The virus changes its face through a process known as drift, which involves small mutations to genes. After a number of alterations the flu genes become virtually unrecognisable to most of the antibodies in the population. The virus also changes by means of a more dramatic process called shift.

It is these shifts - which can include complete changes to proteins such as neuraminidase on the surface of the flu virus - that make a virus unrecognisable, and are associated with major pandemics of flu, such as that in 1918 which killed more than 50 million people.

Drift happens annually, hence vaccines need to be updated each year. This year the UK has been hit by so-called Sydney flu, a mutated form of a flu that emerged in Sydney two years ago.

ÖйúAƬ

What doctors and scientists dread is a sudden shift, the emergence of an unrecognisable virus that could travel - by plane - across the world in a matter of days, without time for a vaccine to be developed and produced. Shift is thought to be the result of two different viral strains mixing in one host cell - say, a human flu virus and a bird one both taking hold in a pig's cell. A hybrid of the human virus could result - which would be unrecognisable to man.

ÖйúAƬ

It is the unpredictability of flu - the not knowing whether a new shift will occur this year or next, or where it will emerge, or from which animal - that makes the prediction of future epidemics so difficult, and a single vaccine ineffectual.

But the Australian scientific breakthrough appears to address this. Colman's research suggested that a single drug could be used on all types of influenza - from a mild mutation to a complete shift.

Colman used X-ray crystallography to take 3D pictures of a protein, neuraminidase, on the surface of different flu viruses. The photographs - that took four years to perfect - magnified the neuraminidase sites 100 million times.

"We found nearly the entire surface of the neuraminidase is variable," explains Colman. "But the bit where neuraminidase was active was the same from one to another. From one day to the next, we suddenly saw the active centre of neuraminidase was conserved across all strains.

"This sowed the seed for a new drug that had an advantage over existing vaccines by working against all flu strains. I didn't know then how big a problem flu was in human health. It was just a neat scientific problem."

Colman, together with colleagues at the Victorian College of Pharmacy, including Mark von Itzstein, with whom he shared the Australia Prize in 1996 - set about designing a blocking drug that would prevent the conserved area of neuraminidase from working.

The drug they developed was initially funded through an Australian start-up company Biota, and eventually - once a drug was proposed - GlaxoWellcome took over bringing it to market.

Even if Relenza does get the Nice thumbs-up in 2000, in the end it might still be flu that has the last laugh. Colman and colleagues have been experimenting to see how flu develops drug resistance to Relenza in lab conditions.

ÖйúAƬ

"We have been looking at what might happen when it is widely used by people," says Colman, who has produced drug-resistant strains in the lab. "We can expect flu will become resistant to anything we offer it," he concludes. But there is a glimmer of hope - lab-produced resistant strains of flu so far appear to be weakened forms of the virus and initial clinical trials have shown only limited development of resistance to Relenza.

Books, page 28; Agony Aunt, page 42