Virions from the H1N1 outbreak in April 2009. Universal vaccines based on highly conserved antigens could protect against this and multiple other influenza strains, but these vaccines are some way away. Credit: CDC/Science Photo Library

The onset of winter in the southern hemisphere could determine whether the novel, swine-derived influenza A (H1N1) strain detected in Mexico in April will evolve into a full-fledged pandemic virus. Health authorities around the globe are by no means fully equipped to cope with a severe influenza pandemic at this stage, given the lead times and capacity constraints associated with current vaccine production processes—as was the case in 2004, when a dangerous new avian strain (H5N1) emerged in Asia. In the past five years, however, public sector laboratories and biotech companies have made considerable progress in developing modern alternatives to the cumbersome, egg-based manufacturing process that large flu vaccine producers have relied on for decades.

One example is OptaFlu, made by Basel-based Novartis, the first product based on a more flexible mammalian cell culture process, which has recently become available in Europe. And in the past two years, according to a report published in February by New York–based consultants Oliver Wyman, potential global manufacturing capacity for pandemic vaccines has increased by 300% (for avian H5N1 strains at least) because of process improvements and dose-sparing strategies adopted by producers. Nevertheless, it will take several more years for innovative vaccines—based on recombinant approaches involving fusion proteins, DNA sequences or virus-like particles (VLPs)—to be available at the scale required to cope with a major pandemic.

As Nature Biotechnology went to press, the Geneva, Switzerland–based World Health Organization (WHO) had not altered its recommendations to the dozen or so large and small vaccine producers that are licensed to make seasonal vaccines based on hemagglutinin (HA) and neuraminidase (NA) viral antigens. “Unless we receive the order to produce H1N1 vaccine as a priority we will not be producing it,” says Albert Garcia, spokesman for Sanofi Pasteur, the vaccines arm of Paris-based Sanofi Aventis, which produced about 170 million of the world's total supply of around 400 million seasonal flu vaccines last year. In May, the company received an FDA license for a second production facility at its Swiftwater, Pennsylvania, location, which will add another 100 million doses when fully up and running.

Production of seasonal, trivalent vaccine for the northern hemisphere's 2009–2010 flu season, based on WHO recommendations issued in February on the three most likely circulating strains, is nearing completion. A decision to move to a monovalent, pandemic vaccine would be based on a combination of the epidemiology and the severity of the virus, which, so far, has largely resulted in relatively mild illness except in Mexico, which for unexplained reasons, has experienced far more deaths than any other country.

In the meantime, Sanofi Pasteur—and other large influenza vaccine makers, such as GlaxoSmithKline (GSK), of London, Novartis and Baxter of Deerfield, Illinois—are gearing up to be ready to produce pandemic vaccine using seed strains propagated at WHO-designated labs. GSK estimates it will take 4 to 6 months to generate the vaccine. In mid-May, the UK Department of Health placed an order of up to 90 million doses of pandemic vaccine from GSK and Baxter, with the former company indicating that the governments of France, Belgium and Finland also intend to purchase 50 million, 12.6 million and 5.3 million doses, respectively.

Many biotech firms have also been developing novel flu vaccines (Table 1), based on alternative production methods. They have begun work on prototype H1N1 vaccines to demonstrate their capabilities—not least to potential large pharma partners, which, for the most part, have been slow to embrace novelty in this market. “There's a huge inertia in the system,” says Alan Shaw, CEO of one such firm, Cranbury, New Jersey–based VaxInnate. That's because existing producers have invested large amounts of time and cash in optimizing their current production processes, which, though far from perfect, are cheap and usually (though not always) reliable.

Table 1 Selected pandemic influenza vaccines in development

VaxInnate, which closed a $30 million financing round in May, is developing both seasonal and pandemic vaccines, based on recombinant proteins that can be rapidly produced in high volumes in bacterial expression systems. A 1,000-liter fermentation process, Shaw says, could produce around 400 million doses in a matter of months. In each instance, VaxInnate is combining a viral antigen with an immunostimulatory bacterial flagellin protein, which binds Toll-like receptor (TLR)5 and triggers both an innate immune response and a more efficient adaptive response against the virus. Its seasonal vaccine incorporates a hemagglutinin (HA) antigen, whereas its pandemic vaccine is based on a more highly conserved—but less immunogenic—antigen, the extracellular domain of the M2 viral matrix protein (M2e). “The real hurdle here is M2 has never been shown to protect humans against disease—it works well in mice,” says Shaw.

Universal vaccines, based on highly conserved viral antigens, such as M2e, could provide multi-year protection against multiple influenza strains (Table 2). They could be stockpiled in advance allowing vaccine makers to get off the annual reformulation treadmill needed to keep up with the HA and NA antigens' mutability. Other recent work has suggested that a concealed hydrophobic pocket in the conserved stem region of HA might also be a conserved epitope suitable for vaccine development (Nat. Struct. Mol. Biol. 16, 265–273, 2009; Science, published online, doi: 10.1126/science.1171491, February 26, 2009).

Table 2 Universal influenza vaccines in development

Several universal vaccines have already entered the clinic, but progress has been slow. “I believe one of the reasons these things have not moved very quickly is the results have not been spectacular,” says Dino Dina, CEO of Dynavax Technologies, of Berkeley, California. Next year, Dynavax aims to start a clinical trial of another candidate universal vaccine, a recombinant protein comprising two conserved viral antigens, nucleoprotein (NP) and M2e, fused to an immunostimulatory sequence that acts as a TLR9 agonist. “Nucleoprotein generates immunity during natural infection, but it's only present in trace amounts in conventional vaccines,” says Dina. The protein, he says, elicits a cytotoxic T-cell response, which could help to reduce viral spread and transmission. “In a pandemic kind of setting that would be a very valuable feature.”

But others see universal vaccines as a long-term bet. “Our biggest concern is that the regulatory pathway for universal vaccines is not clear,” says Rahul Singhvi, CEO of Rockville, Maryland–based Novavax. The firm uses VLP technology to develop vaccines based on HA and NA and the structural protein M1. A baculovirus vector expressed in an insect cell culture system produces particles, which closely resembles the native virus. “To the immune system it appears like there's a natural infection at the site of immunization,” says Singhvi. This approach, he says, would enable large-scale manufacturing within around 12 weeks of a pandemic strain being characterized.

The company is also offering, in conjunction with GE Healthcare, a subsidiary of Fairfield, Connecticut–based GE, a low-cost, portable, disposable manufacturing system for pandemic vaccines. “You can do this in low-infrastructure environments,” Singhvi says.

Quebec-based Medicago is also harnessing VLP technology, but in a radically different setting. The company has developed a transient gene expression system in the plant species Nicotiana benthamiana, a close relative of the tobacco plant, which can produce VLPs comprising the viral HA antigen only. It relies on an Agrobacterium plasmid to deliver the construct to the plant cells. Frederic Ors, Medicago's vice president of business development, says the purified VLPs are highly immunogenic, and the production process is also relatively low cost. “All you need is a greenhouse,” says Ors. “The biomass production is cheap, even in comparison to eggs.”

It will be several years yet, however, before any of these innovations—and others in development at competitor firms—will be ready for commercial rollout. In the meantime, drug therapy will remain a vital frontline defense against a pandemic (Box 1).

At this point, it is not yet clear whether the current pandemic alert will escalate further or will peter out, as recent avian flu epidemics have done. What is certain is that a vaccine for swine-derived H1N1 lies several months away.