Massive amounts of H5N1 vaccine would be needed if there’s a bird flu pandemic. Can we make enough?

The unsettling reality of H5N1 bird flu circulating in dairy cow herds in multiple parts of the United States is raising anxiety levels about whether this dangerous virus, which has haunted the sleep of people who worry about influenza pandemics for more than 20 years, could be on a path to acquiring the ability to easily infect people.

To be clear, there is no evidence that this is currently the case — the sole confirmed human case reported in Texas three weeks ago was in a farm worker who had contact with cattle. There is no way to predict if the virus will acquire the capacity to spread between people, or when and under what conditions it would make that fateful leap if it does.

But the first signs that H5N1 — or any new flu virus — was starting to spread from person to person would trigger a race to produce massive amounts of vaccine to try to mitigate the damage a flu pandemic might be expected to cause. While the 2009 H1N1 pandemic is estimated to have killed about a quarter-million people worldwide — severe flu seasons sometimes kill more — the 1918 Spanish flu pandemic is believed to have killed between 50 million and 100 million people, many times more than Covid-19.

The good news: The world makes a lot of flu vaccine and has been doing it for decades. Regulatory agencies have well-oiled systems to allow manufacturers to update the viruses the vaccines target without having to seek new licenses. The United States even has some H5 vaccine in a stockpile that it believes would offer protection against the version of the H5N1 virus infecting dairy cattle, though there would not be nearly enough doses for the entire country.

The bad news: The current global production capacity isn’t close to adequate to vaccinate a large portion of the world’s population in the first year of a pandemic. And batches of flu vaccine, often (though not always) produced in hen’s eggs, take months to produce.

The Covid pandemic saw vaccine manufacturers around the world pump out unprecedented amounts of vaccine in an extraordinarily short period of time. By February 2022, 14 months after vaccination efforts began, it was estimated that 11 billion doses of vaccine had been produced and delivered. Could that record be matched — broken? — if H5N1 starts a pandemic? Would the profound inequity of the Covid vaccine distribution, which had affluent countries awash with doses while low-income countries waited for supplies, be remedied or repeated?

Experts STAT interviewed suggested that in some respects, the world is better positioned to produce pandemic flu vaccines, if the need arises. But a number warned that assuming the successes of Covid vaccine production would automatically influence the speed and scale of pandemic flu vaccine production would be unwise.

“You just have a different [production] system for flu vaccines than we developed on the fly for Covid,” said Richard Hatchett, CEO of the Coalition for Epidemic Preparedness Innovations, better known as CEPI, an international organization tasked with spurring development of medical countermeasures for diseases that can cause epidemics and pandemics. “And the … system that we developed for Covid can’t just be repurposed for flu.”

Let’s explore some of the ways the world is perhaps better prepared than it was in the past, and some of the snarls that could slow efforts to protect against a dangerous pandemic.

What the need would be

To gain protection against seasonal flu, most people get vaccinated with one dose of vaccine. Young children who’ve never been vaccinated against flu are an exception; kids ages 6 months to 8 years must get two doses of vaccine, at least four weeks apart, the first time they are vaccinated.

Why? Because they are considered to be “naive” to influenza, and their immune systems need to be primed (introduced to flu viruses) and then boosted. Anyone over age 8 is likely to have encountered flu before; they’ve already been primed. One dose is sufficient to acquaint immune systems to the evolved versions of previously encountered seasonal flu viruses.

But with a brand-new flu virus — and H5N1 definitely falls into that category for humans — everyone will be naive. So the assumption of pandemic planners is that everyone would need two doses of vaccine in the first season of vaccinating against H5N1. (During the 2009 H1N1 pandemic, the new virus was distantly related to one that had circulated for much of the previous century, and one dose was adequate for protection. There’s no expectation that would be true with an H5 pandemic.)

That math is daunting: The 8.1 billion people worldwide times two is 16.2 billion doses. Babies under the age of 6 months aren’t vaccinated against flu — their immune systems aren’t yet sufficiently developed for it to be effective — so that number would be a bit lower. But there’s always waste in vaccine delivery, so the 16 billion probably isn’t far off.

Who’s most at risk?

Speaking of kids, there’s a dynamic at play with H5N1 that wasn’t seen in the Covid pandemic, one that adds to the complexity of vaccinating the planet in a flu pandemic.

The vast majority of Covid deaths were in older adults; there were relatively few deaths in children. But flu infections are hardest on older adults and little kids. In the 2009 H1N1 pandemic, it is estimated that nearly 1,300 children under the age of 17 died in the United States, a high toll when compared to a typical flu season. And 2009 was mild, in pandemic terms.

Furthermore, kids amplify the transmission of flu in communities, Hatchett said, noting the fall wave of the 2009 pandemic in the U.S. started in late summer, when southern states went back to school. A really bad pandemic could change public opinion rapidly, but at present it seems unlikely there would be much appetite for closing schools to slow spread of a new flu virus. School closures are one of the most reviled of the containment measures that were used in the Covid pandemic.

“The disease is different. It behaves differently. It has different mortality patterns and different transmission patterns. And we would be in a setup because we’d be coming out of Covid with a Covid mindset and Covid biases that would jeopardize our ability to do the right thing if there were a flu pandemic,” Hatchett said.

So while the United States waited into 2022 before starting to vaccinate children against Covid — and some countries didn’t vaccinate children at all — kids will be in the pool of people who need to be vaccinated when the next flu pandemic hits.

What the current production capacity is

An assessment of market capacity published by the World Health Organization in January estimated that the combined global supply was 1.2 billion trivalent doses of flu vaccine, which target three flu virus strains at a time. During a pandemic, the vaccine made would target a single virus, the new one.

So, in theory, 1.2 billion vaccine doses could become 3.6 billion vaccine doses. More than 85% of those doses are made by seven producers, the WHO analysis said, and over 95% of seasonal vaccine doses are used in high- and upper-middle income countries, which will almost certainly have first dibs on pandemic vaccines. More on this later. Back to the math.

How much vaccine would each person need?

The 3.6 billion doses would be enough for 1.8 billion people. But that’s only if the same amount of vaccine used to protect against seasonal flu viruses would protect against H5N1. And there we run into another problem.

Research done in the mid-2000s found that H5N1 is poorly immunogenic in people; it doesn’t trigger a strong immune response unless it’s administered in large amounts, or is given with a boosting compound known as an adjuvant that broadly stimulates the immune system.

In a seasonal flu shot, each component contains 15 micrograms of antigen or vaccine. In that study, which was published in the New England Journal of Medicine, it took two doses of 90 micrograms — 12 times the amount used to protect against a strain in the seasonal vaccine — to induce what’s thought to be a protective response in just over half of the volunteers. In a world where need will outstrip supply, that’s a wholly unworkable dosage.

In the years since, multiple studies have been conducted to see whether adjuvants could lower the amount of antigen needed and stretch supplies. They do. Some studies have even suggested fractional doses, with an adjuvant, could be effective. For example, CSL Seqirus, which supplies flu vaccine to the U.S. market and has a licensed pandemic H5N1 vaccine in the National Pre-Pandemic Influenza Vaccine Stockpile (NPIVS), showed that 7.5 micrograms, with an adjuvant, would generate what’s thought to be a protective response in a portion of adults. The studies were small, so the range of the estimates is broad, but between 28% and 64% of volunteers ages 18 through 64 and 17% to 57% of volunteers ages 65 and older developed what is thought to be a protective level of antibodies from the 7.5 microgram shot.

But not all flu vaccine manufacturers use or have proprietary adjuvants. CSL Seqirus and GSK do; Sanofi, a major producer of flu vaccines, currently does not. And the global supply of adjuvants could be a bottleneck in any effort to vaccinate the world against H5N1, experts say.

(About two weeks after publication of this story, Sanofi struck an agreement with embattled Covid vaccine manufacturer Novavax that, among other things, gives Sanofi a non-exclusive license for Novavax’s Matrix-M adjuvant.)

“I would be very curious to know what the global production capacity for the relevant adjuvants that we have data for would be. And I suspect it would be vastly insufficient to what is needed,” Hatchett said.

The U.S. has stores of adjuvants in the NPIVS, said David Boucher, director of infectious diseases preparedness and response for the Administration for Strategic Preparedness and Response (ASPR), a division of the Department of Health and Human Services.

How quickly could the U.S. respond?

The United States has been doing active pandemic preparedness work for at least the past two decades. It understood that domestic production of flu vaccine would be critical in a pandemic, when borders might close to exports if a situation became dire. (Witness India’s refusal to export Covid vaccines in spring 2021 when it was hit with a huge wave of cases.) The government incentivized development of flu vaccine manufacturing and fill-and-finish operations in the U.S.

The country, through ASPR, has worked with flu vaccine manufacturers to develop and test batches of H5N1 vaccines targeting different versions of the virus. H5N1 vaccine has already been tested in children, which will help speed availability for a group that is typically at the end of the testing list. (Manufacturers are reluctant to test vaccines in children until they have lots of adult safety data.)

There are currently four types of H5N1 vaccine in the NPIVS, said Boucher. Analysis of the viruses they are based on and the viruses from the cattle outbreaks suggests that two of the stockpiled vaccines are sufficiently close, genetically, to be useful, should this version of the virus take off, the WHO and the CDC have stated.

“We have right now two candidate vaccine viruses that are very well matched to the current virus from those we have,” Boucher said. About 10 million doses are being stored in bulk, and “hundreds of thousands” of doses have already been placed in vials.

That’s a start. But with two doses needed per person, it would not go very far. Boucher said decisions on who the first doses would go to have yet to be made.

Those vaccines haven’t been approved by the Food and Drug Administration, but the agency has a fast-track process for updating flu vaccines, and a lot of recent experience with emergency authorization of pandemic vaccines. It can make this happen fast.

“Some fraction” of the doses in the filled vials could be available within weeks, Boucher said, depending on how quickly the regulatory process went. And while the bulk antigen was being put into vials, large-scale manufacturing would kick into gear. He estimated that tens of millions of doses could be manufactured over a four- or five-month period. An ASPR spokesperson later put the figure at about 125 million doses. (Remember, 125 million doses would be enough to vaccinate 62.5 million people, about a fifth of the U.S. population.)

Kathleen Neuzil, a longtime influenza vaccine researcher at the University of Maryland who takes over as director of the Fogarty International Center at the National Institutes of Health in early May, studied one of the two stockpiled vaccines that Boucher mentioned. It is based on an H5N8 virus and has the same key component as H5N1. (The major target of a flu vaccine is the hemagglutinin protein on the viruses’ surface, the H in a strain’s name.) Her group tested blood from people vaccinated with it against a virus related to the ones that have been seen in cows, which come from a subgroup of H5N1 known as 2.3.4.4b. Although the vaccine targets a 2.3.4.4c virus — a slightly different clade — the researchers saw good antibody responses in people who had received two 15-microgram doses of the vaccine, with adjuvant.

“We had tested this H5N8 vaccine that’s in the stockpile to that strain, and it had pretty good cross reactivity — if you use an adjuvant,” Neuzil told STAT.

Would the stockpiled vaccine protect us, if H5 triggers a pandemic?

Two vaccine and pandemic preparedness experts who used to work for the FDA recently questioned how useful the stockpiled vaccine would be, in an opinion piece in STAT. “Let’s say that by a miracle that it really is a good match, that it’s protective and all that. It’s available in such limited numbers it’s just not meaningful for the response,” Luciana Borio, a former acting chief scientist at FDA and former member of the National Security Council, said in an interview.

Michael Osterholm, director of the Center for Infectious Disease Research and Policy at the University of Minnesota, said it would be prudent to remember the vaccine experience of the 2009 pandemic when making projections about how effective the stockpiled vaccine might be against a virus that evolves to be able to spread from person to person. A study of how well the vaccine worked then showed good protection for young people but no discernible protection for older adults. “You had the strain in the vaccine. And you still saw a relative lack of protection that you would expect to have with a closely matched virus,” he said.

Then there’s the issue of how long it takes to produce additional doses, if they are needed. Making a new type of flu vaccine requires manufacturers to learn how to grow large volumes of the viruses. Some viruses grow better than others; sometimes it takes time to learn how to optimize growth. In 2009, manufacturers initially struggled to produce adequate supply of viruses, delaying rollout of the vaccines.

In fact, though tens of millions of Americans were vaccinated against H1N1 in 2009, the fall wave of that pandemic — when the vast majority of U.S. cases occurred — had already largely subsided by the time vaccine doses were available in substantial numbers.

(Source: The Center for Infectious Disease Research and Policy, the University of Minnesota)

Could mRNA vaccines be a wild card?

During the Covid pandemic, messenger RNA vaccines — which are quicker to make because they don’t require growing supplies of virus — hit the market with incredible speed. But it’s not clear what role mRNA vaccines would play in a flu pandemic, if one were to occur soon.

Both Pfizer and Moderna, the major mRNA players, are developing flu vaccines they hope to combine with Covid shots in a bid to boost flagging sales of the latter and grab a share of the pretty saturated seasonal flu vaccine market. Both have reported that for influenza A — H5N1 belongs to that large flu family — their vaccines appeared to induce immune responses similar to those seen generated by traditional flu vaccines.

Moderna is currently testing an H5N1 vaccine, from the 2.3.4.4b subset of viruses, in people. That trial began last summer.

But the trial’s listing in the Clinicaltrials.gov database is cagey about the dosages Moderna is testing, calling them simply dose number 1, 2 and 3. Christopher Ridley, Moderna’s vice president of communications, said the detail was obscured for “competitive reasons.” Given how likely it is for mRNA vaccines to induce unpleasant side effects, especially at higher doses, and given how poorly immunogenic H5N1 viruses are, finding a sweet spot — an mRNA H5 vaccine that would be both effective and tolerable to take — could take some work.

Hatchett thinks that regulators will want more data before they embrace mRNA vaccines as a solution in a flu pandemic — though he and others have noted that in a true crisis, “all bets are off.”

“My intuition would be that we are still in a vulnerable period when it’s not just an instant solution,” he said of the mRNA option. “I think if we make the proper investments, we could very soon get to the point where we know what the dose needs to be and have mRNA vaccines against avian strains as part of our arsenal and scale it quickly. I just don’t think we’ve worked that through yet.”

Boucher said ASPR is looking to add mRNA H5N1 vaccine to the stockpile, having recently issued a request for proposals from manufacturers. It is still assessing the submissions it generated, he said.

Seth Berkley, who stepped down last summer after 12 years as CEO of Gavi, the Vaccine Alliance, raised other questions about mRNA as an option for a pandemic flu response, noting that with the Covid mRNA vaccines, the protection against any infection waned quickly and the vaccines needed to be stored at ultra-cold temperatures, which made them impractical for parts of the world.

“But if you were to say, we want to make flu vaccines for the world, would that be mRNA technology? Or … wouldn’t you want a vaccine, probably adjuvanted, that would take a little longer to make, but might be inexpensive and easy to produce and give longer immunity?” Berkeley mused. “We don’t know, but those are the questions that will have to get asked, and they didn’t get asked in Covid because of the emergency we were in.”

Will lower-income countries have to wait yet again?

In 2009, the Obama administration announced it would donate 10% of the U.S. stocks of vaccine, as they rolled off the manufacturing line, starting after high-risk people in the country were vaccinated. Hatchett was one of the people tasked with running that program.

Other countries were persuaded to join, and the WHO led an effort to redistribute vaccine to countries that could not buy their way into the vaccine purchasing lines. Eventually, Hatchett said, about 200 million doses were donated, but by the time they were available, the pandemic had subsided. When the program was wound down in October of 2010, about 80 million doses had been distributed.

Fast forward to the Covid pandemic. Nearly 2 billion doses of Covid vaccine were distributed through COVAX, an entity set up by the WHO, Gavi, CEPI, and UNICEF (the United Nation’s Children’s Fund) to try to secure vaccine for countries boxed out of the market. And still, vaccine distribution was hugely inequitable. That fact was part of the inspiration for the ongoing efforts to come up with an international pandemic accord that would try to ensure vaccines, drugs, and other medical essentials are more fairly distributed in the next pandemic. Those efforts have struggled to reach a consensus acceptable to all parties.

People who know the flu vaccine market think the dynamics may not be vastly different in the next pandemic than they were in the last two.

“I would say even today, it’s true that the countries with the manufacturing capacity will be the countries with the early access to these vaccines,” Neuzil cautioned.

An in-depth analysis of global seasonal and pandemic flu vaccine production capacity, led by the WHO and based on 2019 data, showed that low- and middle-income countries made up 38% of the world’s population, but produced 2% of seasonal influenza vaccine. There was no flu vaccine production in Africa, and a Moderna plan to build an mRNA vaccine production facility there was recently shelved, a move the company ascribed to the sharp drop in sales of its Covid vaccine.

Hatchett said a problem that existed in 2009 was the fact that high-income countries that bought seasonal flu vaccine had provisions in their contracts that gave them first or early access to pandemic vaccine, if the need arose. When H1N1 emerged, “all of the vaccine that could be produced in Europe was absorbed,” he said. “And I suspect those contingent purchase contracts are probably still in place, and probably you would have the exact same phenomenon again.”

This article has been updated to note that Sanofi has signed a deal to license an adjuvant for use in its vaccines.