Virus verification in a hen's egg. Credit: Ulrich Baumgarten/Getty

In November 2012, as an early and severe flu season bore down on North America, the news from the US Centers for Disease Control and Prevention (CDC) in Atlanta, Georgia, was reassuring. "So far, this season, most (90%) of the influenza viruses … are well-matched to the 2012–2013 influenza vaccine; this should mean that the vaccine will offer good protection," the agency stated.

Critics say British Columbia's grizzly-bear cull not based on science How the $1,000 genome became reality China's moon rover responds but cannot move

Yet vaccine effectiveness against H3N2, the main flu strain circulating that season, proved to be only 46% in adults aged 18–49, 50% in those aged 50–64, and a dismal 9% in people aged over 65, a vulnerable group. Flu hospitalizations and deaths were the highest in almost a decade. What went wrong?

Research published on 25 March in PLoS ONE1 by Danuta Skowronski from the BC Centre for Disease Control in Vancouver, British Columbia, and colleagues at other Canadian public-health research centres, shows that the H3N2 vaccine strain selected by the World Health Organization (WHO) was indeed well matched to the wild viruses circulating at the time. But the strain sent to vaccine makers — which is first adapted to grow better in the hens' eggs used to produce the vaccine — was mismatched and poorly effective, they find.

The likely cause, the scientists found, is three mutations in the egg-adapted strain that were not present in the WHO strain, resulting in changes at key sites of the haemagglutinin surface protein known to affect antibody responses.

Poor outlook

Moreover, the Canadian researchers found that the same problem probably arose with a revised H3N2 component of the 2013–14 flu vaccine (each season’s vaccine is also designed to ward off H1N1 and influenza B type viruses) and the upcoming 2014–15 vaccine. This suggests that effectiveness against H3N2 may be poor in these seasons’ vaccines, too. 

Such mismatches are unlikely to be one-off incidents, says Michael Osterholm, who heads the University of Minnesota’s Center for Infectious Disease Research and Policy in Minneapolis. “I'm sure this has had to have happened before,” he says. “It's another reason why we need to accelerate efforts to come up with really game-changing flu vaccines.”

Flu viruses are continually evolving, and so seasonal flu vaccine strains are changed from year to year to better match the circulating strains. Any reduction in effectiveness has previously been explained as 'antigenic drift' taking place in the circulating strains, a process in which mutations accumulate in the genes encoding the flu antigens seen by the immune system, producing strains that it no longer recognizes.

But the new findings show that it is important to ensure that the egg-adapted virus is well matched, and also to better understand how particular mutations affect vaccine effectiveness, says Skowronski.

Scrambled egg virus

H3N2 viruses do not grow well in eggs, so in preparing the vaccine, the WHO strain is 'reassorted' with other strains that do. These fast-growing 'seed strains' are used for manufacture (see 'The virus grower').

"There is always a concern about egg-adaptation mutations resulting in alteration of antigenicity of the vaccine seed strain," says Doris Bucher, a microbiologist at New York Medical College in Valhalla, New York, whose lab creates such seed strains. Efforts are made to monitor mutations and select for the least changes, she says. Quality control of the seed strains is done by the WHO Collaborating Centres. "I'm surprised that the antigenic difference did not become obvious much earlier in the process of testing the high growth reassortant before it was provided to the manufacturers," says an official at one flu vaccine company who did not wish to be named.

The WHO briefly alluded to the problem of mismatch in February 2013 when it recommended that the H3N2 strain used in the 2012–13 season should not be used for the 2013–14 season but should be substituted with a different strain.

But the Canadian scientists found that the egg-adapted version of the new vaccine strain is similarly mismatched against 2012–13 wild virus, and data in a WHO report released last month suggest a similar mismatch for 2013–14 wild viruses. The finding, says Bucher, "is clearly of concern for the 2014–15 vaccine effectiveness of the H3N2 component".

The WHO has recommended that the new strain be used in the 2014–15 season. The UN agency had not responded to Nature's questions by this article's deadline.

Skowronski says that the findings should prompt discussion of switching vaccine production from eggs, a 1930s technology, to modern production systems using cultured human cells.

Flu vaccination remains the best tool to protect against flu, says Osterholm, but its overall effectiveness is poor compared to vaccines for other diseases. "We need much improved vaccines than what we have now," he says.