After a Devastating Flu Season, FDA Advisers Unanimously Vote to Target Subclade K in Next Fall's Vaccines

The flu virus known as subclade K arrived too late to be stopped by this season's vaccines. It won't get a second chance.

On March 12, 2026, all seven voting members of the FDA's Vaccines and Related Biological Products Advisory Committee (VRBPAC) unanimously recommended that influenza vaccines for the 2026-2027 season include strains targeting subclade K — the H3N2 variant that has dominated this winter's flu season and contributed to the third-highest cumulative hospitalization rate in over a decade [1][2]. The recommendation followed a similar endorsement from the World Health Organization two weeks earlier [3].

The vote marks an attempt to close the gap that defined the 2025-2026 flu season: a virus that mutated after vaccine strains were already locked in, leaving hundreds of millions of doses partially mismatched against the dominant circulating pathogen.

"It's a little bit of science, a little bit of luck," said Arnold Monto, acting VRBPAC chair, characterizing the perennial challenge of flu vaccine strain selection [1].

The Rise of Subclade K

Subclade K — formally classified as influenza A(H3N2) clade J.2.4.1 — was first identified in European surveillance samples in June 2025, months after vaccine manufacturers had finalized the composition of this season's flu shots based on February 2025 recommendations [4][5]. The variant carries seven new mutations in key positions of its hemagglutinin protein, the surface structure the immune system targets and the basis for vaccine design [6].

By late September, subclade K had begun displacing older H3N2 strains. By December, it was overwhelming: CDC genetic characterization showed approximately 90% of H3N2 viruses belonged to subclade K [7]. As of the most recent surveillance data through March 7, 2026, that figure has risen to 92.6% [8].

The rapid takeover wasn't unexpected in its general contours — H3N2 is the most mutation-prone of the major flu virus subtypes, and antigenic drift has plagued vaccine matching for decades. But the speed and completeness of subclade K's sweep caught many off guard.

"It emerged in October and arrived too late for inclusion in last year's vaccines," CDC influenza expert Lisa Grohskopf told VRBPAC members [1]. The result was a season defined by the mismatch.

A Season of Severe Toll

Source: CDC FluSurv-NET / FluView Week 9

Data as of Mar 7, 2026CSV

The numbers tell a grim story. Through the week ending March 7, 2026, the CDC has recorded 27,242 laboratory-confirmed influenza-associated hospitalizations through its FluSurv-NET surveillance system, yielding a cumulative hospitalization rate of 78.2 per 100,000 population — the third highest since tracking began in the 2010-2011 season [8].

The burden has fallen hardest on the extremes of age. Adults 65 and older have experienced a cumulative hospitalization rate of 256.9 per 100,000, while children under 5 have seen rates of 82.4 per 100,000 [8]. One hundred and one influenza-associated pediatric deaths have been reported this season. Among children who were eligible for vaccination and had known vaccination status, approximately 85% of those deaths occurred in children who were not fully vaccinated [8].

The CDC's broader burden estimates — which account for undertesting and unreported cases — paint an even larger picture: at least 27 million flu illnesses, 350,000 hospitalizations, and 22,000 deaths during the 2025-2026 season [8][9].

At its peak during the week ending December 27, 2025, the weekly hospitalization rate reached 12.8 per 100,000, driving flu activity to levels not seen in nearly three decades [10]. The season has since declined significantly — the most recent weekly rate stood at 1.7 per 100,000 — but the cumulative damage has been extensive.

CDC's in-season severity assessment has classified the season as "moderate" across all age groups, a designation that speaks to the per-case severity rather than the overall burden [8]. Importantly, there is no evidence that subclade K itself causes more severe illness than other H3N2 strains. The toll appears driven by sheer volume of infections facilitated by immune escape from vaccine-induced and prior-infection immunity.

The Mismatch Problem

The 2025-2026 flu vaccine was designed around H3N2 strains circulating in early 2025 — before subclade K existed. That mismatch has measurably reduced vaccine effectiveness, though the picture is more nuanced than early "superflu" headlines suggested.

Interim analyses from multiple countries paint a consistent picture of reduced but still meaningful protection [6][11][12]:

• United Kingdom: 70-75% effectiveness against hospital attendance in children ages 2-17, but only 30-40% in adults [6]
• Canada: Approximately 40% reduction in risk of medically-attended acute respiratory illness due to influenza A(H3N2), including subclade K strains [12]
• France: 36.4% overall effectiveness, ranging from 57.2% in children to 27.7% in adults 65+ [11]
• European multicountry estimates: 52% effectiveness in ages 0-17 and 57% in ages 18-64 [11]

The consistent finding across studies: protection is lower than in well-matched seasons, but the vaccine is not useless — particularly for preventing severe disease in children. For older adults, effectiveness is most attenuated, a finding consistent with the known challenges of generating robust immune responses in aging immune systems.

"This does not mean the vaccine is failing against subclade K," researchers stressed [6]. "The bigger concern is uptake."

A Vaccination Crisis Within the Crisis

Source: GDELT Project

Data as of Mar 14, 2026CSV

The mismatch arrived at the worst possible time. Flu vaccination rates have continued a multi-year decline, hitting their lowest levels in over a decade during the 2025-2026 season. Only 47% of adults and 52% of children received a flu shot — numbers that public health officials describe as deeply concerning [13][14].

The low uptake has compounded the mismatch problem. Even a partially effective vaccine prevents tens of thousands of hospitalizations and thousands of deaths when administered at scale. At under 50% coverage, the population-level benefit is dramatically reduced.

The decline in vaccine confidence is not occurring in a vacuum. It tracks with broader erosion of public trust in health institutions — a phenomenon FDA Commissioner Martin Makary himself has acknowledged. "There's growing public mistrust of government health advice," Makary told NPR in December 2025 [15].

What the VRBPAC Recommended

The committee's March 12 vote endorsed strain recommendations that align with WHO guidance issued on February 27 [1][3].

For egg-based vaccines — still the majority of flu shots manufactured in the United States:

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

For cell culture, recombinant, or nucleic acid-based vaccines:

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

The A/Darwin strains — both the egg-adapted and cell-culture versions — are subclade K viruses, meaning next season's vaccines should match the currently dominant H3N2 variant [1][2]. All three viral components were updated from the current season's formulation, reflecting significant drift across influenza A and B lineages [3].

The recommendations call for a trivalent formulation targeting two influenza A strains (H1N1 and H3N2) and one influenza B strain (Victoria lineage) — the same architecture used in recent seasons following the near-extinction of the B/Yamagata lineage.

The Makary Factor

The final decision on vaccine strain composition rests with FDA Commissioner Martin Makary, not the advisory committee [1]. That fact carries unusual political weight in 2026.

Last year, the FDA under Makary's leadership cancelled the March 2025 VRBPAC meeting that was supposed to recommend strains for the current season's vaccines [16]. The cancellation — announced with no explanation on February 26, 2025 — drew bipartisan criticism. Senators Patty Murray, Tammy Baldwin, and Angela Alsobrooks wrote to Makary raising alarms "over the decision to cancel a critical FDA flu shot meeting amid the worst flu season in 15 years" [17].

During his Senate confirmation hearing, Makary characterized VRBPAC as "merely rubber stamping" WHO decisions — a statement that alarmed vaccine experts who view the committee's independent review as a critical check on the process [18]. The FDA ultimately selected strains for the 2025-2026 season through a closed-door interagency meeting, without input from patient advocates, medical groups, manufacturers, or independent scientists [16].

This year's decision to convene VRBPAC represents a partial restoration of the traditional process, though the committee's reduced membership — just seven voting members, compared to historical norms of 15 or more — has raised questions about whether the panel has been fully reconstituted [2].

The Moderna mRNA Flu Vaccine Standoff

The VRBPAC meeting unfolded against the backdrop of a separate controversy over the FDA's treatment of Moderna's experimental mRNA flu vaccine. In February 2026, the FDA issued a rare "refuse-to-file" letter, declining to even review Moderna's application — a step taken in only about 4% of submissions [19][20].

The FDA claimed that Moderna's Phase III trial, which enrolled more than 40,700 participants, had design flaws because it used a standard-dose flu vaccine (Fluarix) as its comparator rather than a high-dose vaccine for participants over 65 [19]. Moderna fired back, noting that the FDA had agreed to the trial design in writing in April 2024 and that 2 million U.S. seniors still receive standard-dose vaccines annually [20].

After intense public backlash from Moderna and investor pressure, the FDA reversed course on February 18, agreeing to review an amended application with a decision expected by August 5 [21]. The episode further complicated the FDA's relationship with vaccine manufacturers at a moment when trust and cooperation are essential for rapid strain updates.

Manufacturing Against the Clock

Vaccine manufacturers now face approximately six months to produce, test, and distribute updated flu shots before the 2026-2027 season begins in the fall [1]. The timeline is tight but standard — it's the reason strain decisions must be made in February or March each year, even when the currently circulating viruses are still evolving.

The perennial challenge is that flu viruses don't stop mutating while vaccines are being manufactured. Subclade K itself could spawn further drifted variants over the coming months. The A/Darwin candidate vaccine viruses recommended by VRBPAC were specifically selected because they represent the best available match to subclade K, but six months of viral evolution could narrow or widen that match.

For egg-based manufacturing — which accounts for roughly 80% of the U.S. flu vaccine supply — an additional concern is egg adaptation: the mutations viruses acquire when grown in chicken eggs can sometimes alter the hemagglutinin protein in ways that reduce the vaccine's match to circulating strains. This is one reason the committee recommended different specific strains for egg-based versus cell-based and recombinant vaccines [1].

Looking Ahead

The unanimous vote provides a clear scientific signal, but several uncertainties remain.

First, whether subclade K will still be the dominant strain when next flu season begins. H3N2's track record of rapid mutation means the virus could drift further, potentially requiring yet another mid-course adjustment — though the WHO's and FDA's concordance on A/Darwin provides strong confidence in the current recommendation.

Second, whether the decline in vaccination rates can be reversed. The 2025-2026 season provided a stark demonstration of the consequences of low uptake combined with a mismatched vaccine. Public health authorities face the challenge of rebuilding confidence in a vaccination infrastructure that has been subject to political turbulence.

Third, how the FDA's regulatory posture under Makary will affect the broader vaccine ecosystem. The cancelled VRBPAC meeting in 2025, the Moderna refuse-to-file controversy, and the reduced size of the advisory committee all point to an agency in transition — one navigating the tension between the commissioner's stated goal of applying "basic scientific thresholds" and the vaccine community's concern about institutional erosion [15][22].

What is not in question is the scale of the problem subclade K exposed. A virus that emerged months after vaccine formulations were locked in went on to infect an estimated 27 million Americans, hospitalize 350,000, and kill 22,000. The system worked as designed — strain selection committees met, voted, and recommended. Whether the system works fast enough, and whether enough Americans will roll up their sleeves when updated shots arrive, are the open questions heading into fall.