Next season's flu vaccines are being rebuilt around the H3N2 variant from August

The World Health Organization announced on 27 February the strains it recommends for influenza vaccines in the 2026-2027 northern hemisphere season, following a four-day consultation reviewing global surveillance data [s1].

Every component changed. That is the practical consequence of the season the northern hemisphere has just had.

What the season looked like

In August 2025 a notably different variant of A(H3N2) emerged, classified as J.2.4.1 and also referred to as subclade K, and spread rapidly across the globe [s1]. WHO says it contributed to an earlier start to the influenza season in many countries, with several reporting higher-than-usual levels of activity [s1].

Subclade K accounted for the majority of influenza viruses reported across regions [s1]. Influenza A viruses were predominant overall, with other variants of A(H3N2) and A(H1N1) also recorded [s1]. Low levels of influenza B viruses of the B/Victoria lineage were detected, and no B/Yamagata lineage viruses have been recorded since March 2020 [s1].

An antigenically distinct variant emerging in August, ahead of a northern hemisphere season whose vaccine composition was fixed months earlier, is the recurring structural problem in influenza vaccination. Strain selection has to happen far enough in advance for manufacturing; the virus does not observe the deadline.

The recommendations

For egg-based vaccines, WHO recommends [s1]:

• an A/Missouri/11/2025 (H1N1)pdm09-like virus;
• an A/Darwin/1454/2025 (H3N2)-like virus;
• a B/Tokyo/EIS13-175/2025 (B/Victoria lineage)-like virus.

For cell culture-, recombinant protein- or nucleic acid-based vaccines [s1]:

• an A/Missouri/11/2025 (H1N1)pdm09-like virus;
• an A/Darwin/1415/2025 (H3N2)-like virus;
• a B/Pennsylvania/14/2025 (B/Victoria lineage)-like virus.

All six reference viruses are 2025 isolates. The split between the egg-based and non-egg-based lists for the H3N2 and B components reflects a long-standing technical constraint: viruses grown in eggs can acquire adaptive changes that shift their antigenic properties, so the reference strain chosen for egg-based production is not always the same isolate used for cell- or recombinant-based platforms.

These recommendations are used by national regulatory agencies and manufacturers to develop, produce and license vaccines for the season ahead [s1].

The zoonotic half of the meeting

The same consultations assess influenza viruses circulating in animals, particularly those that have infected humans, because of their pandemic potential [s1].

Since 23 September 2025, following the previous consultation, 25 human infections with zoonotic influenza were reported to WHO from six countries [s1]. The majority of these cases had been exposed to infected animals or to environments contaminated with influenza viruses, and no human-to-human transmission was reported [s1].

Experts recommended development of a new candidate vaccine virus for an A(H9N2) virus [s1]. Candidate vaccine viruses are pre-positioned so that manufacturing can start quickly if a pandemic threat emerges [s1] — preparation rather than prediction.

The system doing the work

WHO's twice-yearly consultations bring together experts from WHO Collaborating Centres and Essential Regulatory Laboratories of the Global Influenza Surveillance and Response System, reviewing data generated by GISRS and partners [s1]. GISRS has operated since 1952, making it the longest-standing global platform for systematic disease surveillance [s1].

WHO Director-General Tedros Adhanom Ghebreyesus said the recommendations "rest on the diligent, year-round work" of GISRS and partners [s1].

Globally there are around a billion cases of seasonal influenza annually, including 3–5 million cases of severe illness, and an estimated 290,000 to 650,000 respiratory deaths [s1].

Why the annual reformulation is itself the target

A WHO assessment published nine days earlier makes the case that the current model is the constraint rather than the solution.

Existing influenza vaccines reduce disease burden, but their effectiveness varies by season, product and population group, and protection is limited to a single season [s2]. While 143 countries reported having seasonal influenza vaccines available, upper-middle and high-income countries use the majority of doses [s2].

WHO's Full Value of Improved Influenza Vaccine Assessment estimates that if improved, next-generation or universal influenza vaccines were available and widely used between 2025 and 2050, they could prevent up to 18 billion cases of influenza and save up to 6.2 million lives globally, particularly among older adults, young children and pregnant women [s2].

The assessment also links vaccination to antimicrobial resistance: current influenza vaccine use is estimated to reduce unnecessary antibiotic use by 10 million doses a year, and next-generation vaccines could avert up to 1.3 billion defined daily doses of antibiotics between 2025 and 2050 [s2].

Philipp Lambach, WHO technical lead of the project, said the assessment gives those working on influenza vaccine investment, policy and research priorities "a common set of evidence to catalyze vaccine development" [s2].

Those are modelled projections contingent on vaccines that do not yet exist, and WHO says impact will depend on national disease burden, health system capacity, vaccine pricing and programmatic factors, with safety, efficacy, duration of protection, temperature stability and shelf-life influencing adoption in low- and middle-income countries [s2]. In December 2025 WHO released updated preferred product characteristics calling for vaccines that protect beyond a single season, protect better against severe disease, suit use in low- and middle-income countries, and use technologies transferable to manufacturers in those countries [s2].

What to watch

Whether the updated H3N2 components track subclade K's continued evolution through the northern hemisphere summer. A strain selected in February can drift before September.

Whether zoonotic case counts and the species involved shift before the next consultation. Twenty-five human infections across six countries with no human-to-human transmission is the reassuring version of that number.

And whether the H9N2 candidate vaccine virus is developed on the timeline recommended. Pre-positioned candidates are only useful if they exist before they are needed.

Sources

• [s1] World Health Organization, "Recommendations for influenza vaccine composition for the 2026-2027 northern hemisphere season," 27 February 2026. https://www.who.int/news/item/27-02-2026-recommendations-for-influenza-vaccine-composition-for-the-2026-2027-northern-hemisphere-season
• [s2] World Health Organization, "Next-generation influenza vaccines could save millions of lives, finds WHO," 18 February 2026. https://www.who.int/news/item/18-02-2026-next-generation-influenza-vaccines-could-save-millions-of-lives--finds-who