Faced with the prospect of an influenza pandemic, the World Health Organization (WHO) is weighing up its options for advising manufacturers and governments on developing vaccines. With current manufacturing capabilities, there will be enough vaccine for only a fraction of the world's population, and not before six months from now. And most of that will go to rich countries.

Experts are meeting at the WHO on 14 May to discuss options for proceeding with a vaccine for the currently circulating swine-associated H1N1 strain. One controversial idea being floated is to use a live attenuated vaccine, which could boost the number of doses available from existing plants by 50- to 100-fold.

Manufacturers are lukewarm to the idea. The ordinary seasonal flu vaccine uses inactivated virus, and serious regulatory barriers exist to introducing a live-virus vaccine. Demonstrating efficacy and getting regulatory approval in time would pose "quite significant difficulties", says George Kemble, vice-president of vaccine research and development at MedImmune in Gaithersburg, Maryland, which makes live flu- virus vaccine.

But some experts say the live-virus idea should be entertained. Two factors largely determine whether a vaccine can protect large numbers of people during a pandemic. One is the delay before substantial quantities of vaccine become available — usually around six months, due to the time required to grow the virus in hens' eggs. (Cell culture and other technologies could be faster, but they are not yet ready for prime time.) The second limiting factor is production capacity, currently at around 700 million to 900 million doses of seasonal flu vaccine annually.

Although still limited, production capacity is much better than five years ago — when it was around 300 million — mainly because of measures taken by governments to prepare for a pandemic threat. The seasonal vaccine contains antigens against three circulating flu strains. Switching to producing a single vaccine against just the new H1N1 virus could, in principle, mean that existing plants could make three times as many doses.

Culturing vaccines in chicken eggs is the main time delay in production. Credit: T. Lohnes/AFP/Getty Images

But even if global facilities switched entirely to producing an inactivated H1N1 vaccine, only about 1 billion doses at most are expected to be available by the end of the year, around the time of the Northern Hemisphere flu season. More­over, because the population has little to no pre-existing immunity, the vaccine will probably need to be given in two doses — reducing the actual number of vaccines to 500 million.

Switching even part of the production to a live-virus vaccine would effectively increase production capacity. The virus in such a vaccine is capable of reproducing in humans, so much lower doses can be given. Live-virus vaccines also don't require adjuvants to bolster their effectiveness, can be administered nasally — avoiding the need for syringes — and are thought to provoke a broader and stronger immune response than inactivated vaccine. And whereas one egg yields one dose of inactive vaccine, for a live vaccine it yields somewhere from 50 to 100 doses.

Whatever routes are taken, production capacity will also depend on how well the new H1N1 virus can be grown and cultivated. The news here seems to be good. "We've done 7 cycles, 42 hours each, and it's going very well," Doris Bucher, an immunologist at New York Medical College, told Nature. The US Centers for Disease Control and Prevention in Atlanta, Georgia, asked her to help grow the first reference strains to be sent to manufacturers. The immune response produced by the resultant seed vaccines will need to be tested in clinical trials; if it were, for example, to require three times as much antigen as seasonal flu to prompt an adequate immune response, that would cut theoretical production capacity to a third.

To grow live attenuated vaccines, scientists would reassort the new flu strain with a 25°C cold-adapted strain, which will multiply in the nose but not grow in the higher temperatures of the lower respiratory tract.

Everybody is anxious to have enough seasonal vaccine. ,

At present, only two groups have the technology to produce live attenuated flu vaccines: MedImmune, and Nobilon, a subsidiary of Schering-Plough, which has licensed technology developed at the Institute of Experimental Medicine in St Petersburg, Russia. MedImmune's FluMist is approved for use in the United States for those aged 2-49 years old; older people have been exposed to past pandemic viruses, and their immune systems therefore kill the live vaccine for current circulating strains. The WHO has obtained a licence from Nobilon to allow manufacturers in developing countries to use the Russian technology.

Inactivated vaccine makers seem sceptical about using live attenuated vaccines more widely, even in a pandemic situation. Changing over to a live vaccine would mean introducing new production methods and possibly having to license outside technology, such as that from MedImmune. There are also safety and liability issues, because it would be difficult to organize clinical trials of an untested vaccine quickly enough.

"Given the potential level of global needs, particularly if the threat worsens, all serious vaccine candidates and approaches should be actively considered and vaccine candidates made ready, including live attenuated vaccine," says Jesse Goodman, acting chief scientist of the US Food and Drug Administration. He notes that live attenuated vaccine is approved in the United States for children and young adults, who "may be at particular risk of infection" in the current H1N1 outbreak. Safety, however, is paramount, he adds: "It is also important to keep in mind, even in the face of a pandemic threat, the importance of doing all that is appropriate and possible to assure high vaccine quality and safety, particularly if and as new facilities, processes and products may be considered."

Indeed, memories are still vivid of the 1976 flu-vaccine fiasco. That year, a new swine flu emerged at an army barracks in New Jersey, killing one person but failing to spread further. A mass vaccination campaign ordered by president Gerald Ford caused neurological side effects in some people, and killed 25.

David Fedson, a pandemic-vaccine expert and retired former medical director of French vaccine development company Aventis-Pasteur, now known as Sanofi Pasteur, says that companies are far more comfortable working with inactivated viruses. But he argues that the live-virus approach is the way to go now with H1N1. "Clinical trials of live vaccine have shown it to be safe in people up to 50 years of age," he says. "What we're talking about here is a monovalent, and hence simpler, vaccine with a new H1N1 virus replacing an older H1N1 virus."

boxed-text

Click for larger image

Whatever the WHO decides to do, it will have to balance a switch from a trivalent seasonal flu to a monovalent pandemic vaccine, with the need to make and distribute enough seasonal vaccine in time for the next flu seasons in both hemispheres (see graphic). The WHO is currently surveying all vaccine manufacturers — the bulk of whom are based in Europe — to ascertain where they are in that process, and how ready they would be to switch to a monovalent vaccine.

That they will at some point is a foregone conclusion. Initial information suggests that companies are well along in producing the seasonal H1N1 and H3N2 strains for the Northern Hemisphere, but are having difficulty growing the third influenza B strain. One option may be to go with just the two antigens under way, and drop the B from next year's vaccine. Northern Hemisphere production would then be freed up faster to work on a swine-flu vaccine.

Another question is how to dovetail that with the needs of the Southern Hemisphere, where vaccine production typically starts around November and continues through March. The amounts of vaccine ordered by Southern Hemisphere countries is much lower than that by the north, meaning that northern manufacturers might then have extra time to work on a swine-flu vaccine. A decision on whether to curtail Southern Hemisphere production does not need to be made for weeks or even months, says Marie-Paule Kieny, head of the WHO's Initiative for Vaccine Research.

The WHO is not yet asking its advisory panel to consider calling a halt to seasonal flu-vaccine production. What they are likely to recommend this week is whether or not companies should go to commercial-scale production of a monovalent pandemic vaccine at the earliest opportunity. "Everybody is anxious to have enough seasonal vaccine," says Kieny. One scenario, she says, is that full-scale manufacturing of swine-flu vaccine could start by July at the earliest.

Margaret Chan, director-general of the WHO, will meet on 19 May with the heads of flu-vaccine companies to discuss ways forward, and how developing countries could access the vaccine. Kieny says that experience with mass vaccination campaigns in developing countries, for example for meningitis, should make it easy to quickly deploy a vaccine.

In the meantime, the new H1N1 strain remains susceptible to the antiviral drugs oseltamivir (Tamiflu) and zanamivir (Relenza). Cheaper and more widely available antibiotics and anti-inflammatories, such as statins, could also have a role in limiting mortality in a severe pandemic. "We should be giving attention to looking at new ways of treatment," says Kieny. "We must not give people the impression that if they don't get a vaccine or Tamiflu that they will die." *?

For Nature 's ongoing coverage of the H1N1 outbreak, see www.nature.com/swineflu.