In the spring of 2013 a strain of influenza virus that had never infected humans before began to make people in China extremely ill. Although the virus, known as H7N9, had evolved among birds, it had mutated in a way that allowed it to spread to men, women and children. Within several months H7N9 sickened 135 individuals, of whom 44 died, before subsiding with the advance of summer weather.

We got lucky with H7N9. Had it triggered a pandemic—an explosion of infectious disease across a large geographical area—we would have been woefully unprepared, and millions might have died. The trouble is that every new virus requires a new vaccine, and making new vaccines takes time. Even a typical flu season is brimming with slightly mutated versions of familiar viruses. In most cases, manufacturers anticipate these changes and tweak existing formulas so that they will still work against the new strains. When a virus like H7N9 makes a surprise appearance in people, however, manufacturers must scramble to concoct an entirely new vaccine from scratch, which takes too long to prevent a large number of people from becoming sick and dying.

Public health officials have longed for years to turn the tables, envisioning a “universal” flu vaccine that would be ready and waiting on the shelves to defeat either a marginally mutated strain or a completely unexpected virus. After numerous disappointments, a handful of recent studies indicate that a universal vaccine may at last be close at hand. In an interview last summer National Institutes of Health director Francis Collins suggested that one might be achieved in the laboratory in just five years. Before such a vaccine can reach the general public, however, researchers will have to convince either manufacturers or the government to pay for more studies and demonstrate to the U.S. Food and Drug Administration that the new vaccines are just as safe as those we already use.

Stalking a Killer
Flu vaccines have worked on the same principles since investigators first made them in the 1940s. Each vaccine contains flu antigens—bits of viral molecules that can trigger an immune response. The antigens used in routine flu vaccines are fragments of a mushroom-shaped protein, called a hemagglutinin, that protrudes from a flu virus's surface and helps the pathogen cling to cells inside an infected individual. Once exposed to those bits of protein, a person's immune system produces sentinel molecules called antibodies that will recognize any flu virus possessing the same hemagglutinin and direct an attack against it.

Flu is a rapidly evolving virus, however, and the structure of hemagglutinin in a given strain changes in small ways every season. Even a minor alteration can make it much more difficult for the immune system to identify and eliminate a flu virus that is nearly identical to its earlier version. This is why we have to get new flu shots every year.

Scientists have searched for decades for a way to outsmart the flu virus rather than always hurrying to outpace it. The first glimpse of more efficient vaccines appeared in 1993, when Japanese researchers discovered that mice sometimes generate a single antibody that blocks infection by two flu strains with different hemagglutinins. Fifteen years later several different teams demonstrated that humans occasionally make these cross-protective, or broadly neutralizing, antibodies as well. Most of these antibodies bind not to a hemagglutinin's mushroom cap but rather to its slender stem—a region of the molecule where, as it turns out, less structural mutation takes place. Because the stem's makeup is similar across many strains of flu, the researchers reasoned, an antibody that recognizes it could potentially protect against a range of viral strains with distinct caps.

Building on this discovery, several groups have altered the structure of hemagglutinins, creating a cap to which the immune system does not react. Animals exposed to these tweaked proteins produce cross-protective antibodies that bind to the stalk rather than strain-specific antibodies that home in on the cap. Other scientists are trying to get animals and people to make antibodies against a different viral protein, M2, which is embedded in the flu virus's membrane and helps it enter cells. Like the hemagglutinin stalk, M2 changes little.

Additional teams are focusing on completely different strategies, such as designing a vaccine that encourages the production of T cells, the attack dogs of the immune system. T cells produce broader, longer-lasting immunity than antibodies, but classic flu vaccine formulas do not encourage their activity. Others are administering a sequence of vaccines against different flu strains so that the immune system assembles a diverse antibody artillery.

Much of this research has happened only in the past five years. In fact, for 15 years after the earliest studies in Japan, work on a universal flu vaccine accumulated in mere dribs and drabs—until a pandemic, which killed more young and middle-aged adults than usual, jolted scientists into a higher gear. In April 2009 a highly infectious new strain of swine virus dubbed H1N1 jumped suddenly from pigs to people. Manufacturers had already spent months preparing the vaccine for the 2009–2010 season, which was still a ways away—and that vaccine was useless against the new strain. They had to go back to square one.

Beginning work on the H1N1 vaccine so late in the manufacturing cycle, combined with some peculiarities of the virus—it was not easy to replicate en masse in the lab, which slowed down production—resulted in millions of doses arriving on the market months after planners hoped. By the next spring, H1N1 killed as many as 18,000 people in the U.S. These delays spurred some incremental changes in flu vaccine manufacturing. Yet they also underscored the fact that better techniques cannot solve the root problem of having to rapidly fabricate a new vaccine every time a completely new virus appears.

“We realized that despite all the technology we have, it is very hard to manufacture and deliver a [brand-new] vaccine in time to actually have an impact,” says Kanta Subbarao, chief of the emerging respiratory viruses section at the NIH.

Final Hurdles
Even if researchers who are working on a universal flu vaccine finally overcome all their remaining technical challenges, the real hurdle may be securing both funding for future studies and federal approval for a new product. Asked what he needs to begin trials with people, Peter Palese, who is a professor and chair of microbiology at Mount Sinai Hospital in New York City, laughs and replies, “Money.” Federal or private money? “Any money,” he says.

His answer captures the paradox of research into new flu vaccines. Although current vaccines are flawed and require a lot of time to tweak, they confer some protection most of the time. “Why expend the effort to invest hundreds of millions of dollars to get to something new?” says Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota, which published a lengthy 2012 report scrutinizing the lack of private and government funding for “game-changing” flu vaccines.

Certain unique properties of the most promising universal flu vaccines in production may be a source of additional obstacles. Studies have suggested that the experimental universal flu shots do not trigger as strong an immune response as older vaccines do. Guaranteeing that the new vaccines are as effective as the old ones may mean adding more ingredients or finding new ways to administer them.

Any new flu vaccine is practically guaranteed lengthy FDA examination. Current seasonal vaccines change so little from year to year that they move through FDA review quickly. But a universal vaccine—using new antigens and a new delivery system—would undergo extensive inspection for both efficacy and safety. For comparison, approval of the vaccine Prevnar, the first to confer protection against pneumonia in infants and young children, took 15 years and required very large clinical trials. “There are 60, 70 years” of FDA approvals and clinical experience behind existing flu vaccines, Palese points out. “But if you go in with a new approach, then the FDA will be starting from zero as well.”

Given legally mandated caution on the FDA's side and a natural inclination on the part of manufacturers to stick with a “good enough” product, many have wondered whether a universal flu vaccine will ever reach the market. The emerging consensus seems to be that novel partnerships—in which, perhaps, industry brings the innovation, but government provides the funding—may be able to mitigate the weaknesses on each side. A joint government-industry conference, hosted in 2012 by the FDA and the NIH, concluded that such collaborations offer the best way forward.

“The science is coming along very fast, but we need to figure out how to get to the next step of development,” Subbarao says. Given how quickly the flu virus can mutate—and how suddenly a new lethal virus can leap from animals to people—they had better figure it out fast.