How Is the Flu Vaccine Made? The Production Process

Flu vaccine production is a months-long process that begins with global virus surveillance and ends with hundreds of millions of doses shipped under tight timelines. Most flu vaccines are still grown in fertilized chicken eggs, a method dating back to the 1940s, though cell-based and recombinant approaches are steadily gaining ground. What makes this process unusual compared to other vaccines is that it happens on a near-annual cycle, with new strain recommendations each season and a manufacturing window that leaves almost no room for error.

Picking the Strains Months in Advance

The production process starts not in a factory but in a network of laboratories spread across every continent. Laboratories contributing to the World Health Organization’s Global Influenza Surveillance and Response System monitor circulating viruses year-round, tracking how those viruses change their surface proteins over time.¹ Because influenza viruses constantly mutate, the vaccine composition has to be reviewed before every flu season. For the Northern Hemisphere, WHO typically announces its strain recommendations in February; for the Southern Hemisphere, that announcement comes in September.

The challenge is that vaccine strains must be selected in anticipation of what will circulate months later, because production takes so long.¹ If a new variant emerges after the strain recommendation is finalized, there usually is not enough time to switch. This built-in lag is one reason flu vaccines sometimes miss the mark in a given season.

Growing the Virus in Eggs

Once strains are selected, most manufacturers begin the egg-based process. Candidate vaccine viruses, typically reassortant strains engineered to grow well in eggs, are injected into the allantoic cavity of fertilized chicken eggs. The eggs are incubated for several days to let the virus replicate, and then the virus-containing fluid is harvested. At commercial scale this requires an enormous supply of specially produced, pathogen-free eggs, often numbering in the hundreds of millions per season.

The egg-based method has decades of optimization behind it and remains the workhorse of global flu vaccine supply. But it comes with limitations. Egg supply chains are rigid and can be disrupted by poultry disease outbreaks. And the method introduces a biological problem that researchers have spent years trying to solve: egg adaptation.

Why Growing Flu in Eggs Can Backfire

Influenza viruses sometimes pick up mutations as they adapt to growing in chicken cells rather than human cells. These egg-adaptive mutations in the hemagglutinin protein can alter the virus’s surface appearance, so the vaccine trains your immune system against a slightly different target than the one actually circulating in people.² Some mutations barely matter, while others meaningfully change how well the vaccine matches real-world strains.

Expert estimates suggest egg adaptation reduces vaccine effectiveness by roughly 4 to 16 percentage points, depending on the virus subtype and the age group.³ The H3N2 subtype is the worst offender, especially in younger adults. That is a meaningful hit, roughly comparable to the reduction caused by antigenic drift itself, where the circulating virus has simply evolved away from the vaccine strain over the course of the season.³ This is one of the central frustrations of flu vaccine manufacturing: even when you pick the right strain, the production method itself can erode the match.

Cell-Based Production

Cell-based flu vaccines use mammalian cell lines instead of eggs. The virus is grown in cultured cells, harvested, and then processed much like an egg-grown vaccine. Because the virus replicates in mammalian cells, it is less likely to pick up the egg-adaptive mutations that can weaken an egg-based vaccine’s match to circulating strains.

Optimizing cell-based manufacturing has been an active area of research. Process parameters like the timing of infection, the amount of virus used to seed each batch, and the harvest window all affect how much usable virus you get out. Very low initial doses of virus, for instance, have been shown to produce much higher final yields than heavier seeding.⁴ Several cell lines have been developed for this purpose, including Madin-Darby canine kidney (MDCK) cells and the human cell line PER.C6, which was shown to efficiently propagate both influenza A and B viruses.⁵

From a pandemic-preparedness standpoint, cell-based production offers a critical advantage: it does not depend on a supply of pathogen-free eggs, which can become scarce during an avian influenza outbreak, exactly when you need to ramp up vaccine production the fastest.⁶ Cell culture facilities can also be scaled up more flexibly than egg-production farms.

Recombinant Vaccines and the Egg-Free Approach

Recombinant flu vaccines take a completely different route. Instead of growing live influenza virus at all, manufacturers take the gene for the hemagglutinin protein from the recommended vaccine strain, insert it into a baculovirus vector, and use insect cells to churn out large quantities of the protein. The purified hemagglutinin is then formulated directly into the vaccine.⁷ The best-known product made this way is Flublok.

The key advantage here is that you never need to handle live influenza virus, and you completely sidestep egg adaptation. The manufacturing platform is essentially a “plug and play” system: once you have the gene sequence for the target protein, you can swap in a new strain’s gene without overhauling the production line.⁸ That flexibility is attractive not only for seasonal flu but also for pandemic response, when speed matters enormously. Large-scale mammalian cell culture facilities that were originally built for manufacturing monoclonal antibodies could potentially be repurposed for this kind of vaccine production as well.⁸

Killing and Splitting the Virus

For egg-based and cell-based vaccines, once the virus is harvested, it needs to be inactivated, meaning rendered unable to cause infection. The two main chemicals used for this are formaldehyde and beta-propiolactone (BPL).⁹ Both get the job done, but they work differently at the molecular level. Formaldehyde cross-links proteins, while BPL chemically modifies the virus’s genetic material. The choice of inactivating agent can affect the structure of the final vaccine product.

After inactivation, most flu vaccines are “split,” meaning the virus particles are broken apart using detergents. Common splitting agents include Triton X-100 and diethyl ether, used alongside surfactants like Tween.¹⁰ The split vaccine contains fragments of viral proteins rather than intact virus particles, which tends to produce fewer side effects while still generating an immune response. After splitting, the detergent itself has to be removed through chromatography or evaporation before the product moves to formulation. Some vaccines are purified further to contain only the surface proteins, creating what are called subunit vaccines.

Testing Potency Before It Leaves the Plant

Before any batch ships, manufacturers must verify that it contains enough active hemagglutinin protein to trigger an immune response. The standard method for decades has been the single radial immunodiffusion (SRID) assay, which measures how much HA antigen is present in a sample by diffusing it through a gel embedded with specific antibodies.¹¹

Researchers have explored faster, more modern alternatives like high-performance liquid chromatography and gel-based protein analysis. But these newer methods have a known limitation: in direct comparisons, they failed to detect losses of potency that SRID picked up, and their readings did not reflect what actually happened in animal studies of immune protection.¹¹ In other words, a batch could look fine by the newer tests but produce a weaker immune response in practice. Until a replacement proves it can catch what SRID catches, the decades-old gel diffusion method remains the gold standard.

This creates a practical bottleneck. The SRID assay requires strain-specific reference reagents, antibodies and antigen standards that have to be prepared fresh each season after the new vaccine strains are announced. Developing those reagents takes weeks, and manufacturers cannot release finished doses until the reagents are ready. It is one of the less visible chokepoints in the annual production timeline.

Adjuvants and Why Some Flu Shots Have Them

Most standard-dose flu shots for healthy younger adults contain no adjuvant; they rely entirely on the viral protein to stimulate an immune response. But certain populations, especially older adults, tend to produce a weaker immune response to plain flu vaccine. Adjuvants are ingredients added to boost that response.

The best-known flu vaccine adjuvant is MF59, an oil-in-water emulsion used in the vaccine marketed as Fluad for people 65 and older. After injection, MF59 attracts immune cells to the injection site, creating a localized inflammatory environment that amplifies the body’s reaction to the vaccine antigen. The antibodies produced tend to bind more tightly to their target than those generated without the adjuvant.¹² More broadly, adjuvants enhance both antibody and T-cell responses, which is a meaningful advantage in people whose immune systems are less responsive due to age.¹³

Adjuvanted vaccines do tend to cause more soreness and redness at the injection site, a trade-off of the stronger local immune response they are designed to produce.

How the Nasal Spray Vaccine Is Made Differently

The nasal spray flu vaccine (FluMist) takes a fundamentally different manufacturing approach. Instead of an inactivated virus, it uses live viruses that have been cold-adapted, meaning they can replicate in the cooler temperatures of the nasal passages but not in the warmer environment of the lungs. These cold-adapted, temperature-sensitive properties come from a set of specific mutations originally developed through serial passage of a master donor virus at low temperatures decades ago.¹⁴

To make each season’s nasal spray, the surface protein genes from the recommended circulating strains are combined with the internal genes from the master donor virus using reassortment. This gives you a virus with the right surface identity to match circulating flu but the attenuated internal machinery that prevents it from causing real disease. A key manufacturing concern is genetic stability: if the attenuating mutations reverted during large-scale production, you could end up with a vaccine that behaves more like a wild virus. Studies examining full genome sequences at multiple manufacturing stages have confirmed that the critical cold-adapted, temperature-sensitive, and attenuated mutations remain stable throughout production.¹⁵

mRNA Flu Vaccines on the Horizon

The success of mRNA COVID-19 vaccines has pushed several companies to develop mRNA-based flu vaccines, some of which are now in late-stage clinical trials. The basic concept is familiar from the pandemic: instead of growing virus or producing viral protein in a cell factory, you synthesize mRNA encoding the target protein (usually hemagglutinin), package it in lipid nanoparticles, and inject it. Your own cells then make the protein and present it to the immune system.

For flu vaccine manufacturing, mRNA has some appealing properties. The production process is largely synthetic and can be adapted to new strains quickly, since all you need to change is the mRNA sequence. Scalable manufacturing processes for mRNA have been developed that are adaptable to the regulatory standards required for clinical use, and products made at manufacturing scale have shown good immune responses in animal models.¹⁶ The big open question is how mRNA flu vaccines will perform head-to-head against existing options in large human trials, and whether the platform can match the cost efficiency of established egg-based production.

Regulatory Approval Without New Clinical Trials Each Year

A reasonable question is whether each year’s flu vaccine goes through a full approval process. The answer is no. Once a flu vaccine product has been licensed through a full regulatory review, annual strain updates are handled through supplement filings. Manufacturers submit the updated strain composition and manufacturing data, but they do not have to conduct new clinical trials for each year’s inactivated or recombinant vaccine just because the strains changed.¹⁷ Regulatory agencies treat the strain change as a minor update to an already-proven product and manufacturing process, not as a new product requiring a fresh efficacy trial.

This streamlined pathway is what makes the annual vaccine cycle possible at all. If every strain change required clinical trials enrolling thousands of volunteers, the timeline would blow past the start of flu season before a single dose could ship.

Pandemic Manufacturing and Its Bottlenecks

Seasonal flu vaccine production operates on a tight but predictable annual rhythm. Pandemic vaccine production throws that rhythm out the window. When a novel influenza virus emerges with pandemic potential, like H5N1 avian flu, the entire pipeline has to start from scratch with a strain that may grow poorly in eggs and for which no reference reagents exist.

The egg-based system is especially vulnerable in pandemic scenarios because a poultry-origin influenza outbreak can simultaneously threaten the egg supply itself.⁶ Cell-based and recombinant platforms were developed in part to address this weakness, and both offer faster and more flexible responses to pandemic threats.⁶ Insect-cell-based recombinant systems, for instance, can begin producing protein as soon as the target gene sequence is available, without needing to grow live virus. Still, all current platforms face some combination of limitations in scalability, speed, or cost when demand surges suddenly.¹⁸

The Push Toward a Universal Flu Vaccine

The entire annual manufacturing cycle exists because influenza’s surface proteins keep changing. If a vaccine could target a part of the virus that stays the same across strains and seasons, you would not need to re-formulate and re-produce the vaccine every year. That is the idea behind universal flu vaccine research.

The most advanced approach focuses on the stalk domain of the hemagglutinin protein. The globular head of HA, which current vaccines target, is the region that mutates rapidly. The stalk, by contrast, is highly conserved across influenza subtypes, meaning it changes much less from strain to strain.¹⁹ One strategy uses chimeric hemagglutinins, engineered proteins that pair unfamiliar head domains with a consistent stalk, to redirect the immune system’s attention toward the conserved region. This chimeric HA approach is already in clinical trials.¹⁹

Another line of work has produced “headless” HA immunogens, proteins consisting only of the conserved stalk with the variable head removed entirely. Early results suggested these constructs can generate immune responses potent enough to justify further development.²⁰ If any of these approaches eventually works in large human trials, it would not just change how often you need a flu shot. It would reshape the manufacturing process itself, eliminating the annual scramble to select, grow, and test new strains against an ever-moving target.

1
PubMed Central. Models for predicting the evolution of influenza to inform vaccine strain selection
2
PLoS Pathogens. Egg-adaptive mutations of human influenza H3N2 virus are contingent on natural evolution
3
PubMed Central. Estimation of Reduction in Influenza Vaccine Effectiveness Due to Egg-Adaptation Changes—Systematic Literature Review and Expert Consensus
4
PubMed. Bioprocess optimization for cell culture based influenza vaccine production
5
PubMed. The human cell line PER.C6 as a cell line for influenza virus production
6
PubMed Central. Current and emerging cell culture manufacturing technologies for influenza vaccines
7
PubMed Central. Production of a novel influenza vaccine using insect cells: protection against drifted strains
8
PubMed Central. Recombinant protein vaccines produced in insect cells
9
PubMed Central. Inactivated or damaged? Comparing the effect of inactivation methods on influenza virions to optimize vaccine production
10
PubMed Central. Influenza Vaccine Manufacturing: Effect of Inactivation, Splitting and Site of Manufacturing. Comparison of Influenza Vaccine Production Processes
11
PubMed. Comparison of single radial immunodiffusion, SDS-PAGE and HPLC potency assays for inactivated influenza vaccines shows differences in ability to predict immunogenicity of haemagglutinin antigen
12
Infection & Chemotherapy. Fluad®-MF59®-Adjuvanted Influenza Vaccine in Older Adults – Section: MF59: mechanism of action
13
PubMed. Immunologic correlates of protection and potential role for adjuvants to improve influenza vaccines in older adults
14
PubMed Central. Reversion of Cold-Adapted Live Attenuated Influenza Vaccine into a Pathogenic Virus
15
PubMed. Genetic stability of live, cold-adapted influenza virus components of the FluMist/CAIV-T vaccine throughout the manufacturing process
16
PubMed Central. Development of mRNA manufacturing for vaccines and therapeutics: mRNA platform requirements and development of a scalable production process to support early phase clinical trials
17
PubMed Central. An overview of the regulation of influenza vaccines in the United States – Section: Annual update of seasonal influenza vaccines
18
Open Access Journal of Microbiology & Biotechnology. Silkworm-Based Vaccine Production for H5N1: A One Health Approach to Pandemic Preparedness
19
PubMed Central. Universal Influenza Virus Vaccines That Target the Conserved Hemagglutinin Stalk and Conserved Sites in the Head Domain
20
PubMed Central. Influenza virus vaccine based on the conserved hemagglutinin stalk domain