FDA Data Requirements for Strain-Change and Variant Updates to Influenza and COVID-19 Vaccines

FDA data requirements for strain-change and variant updates to influenza and COVID-19 vaccines

FDA handles influenza and COVID-19 vaccine composition updates through supplements to an existing biologics license application (BLA). The data each update needs depends on how much manufacturing and clinical experience supports predicting the immune response to a new strain. Seasonal influenza is the established model: composition is selected each year, manufacturers file strain-change supplements, and CBER approves them without new clinical trials. COVID-19 vaccines moved toward that model between 2022 and 2024. They started with clinical immunobridging, then moved to approvals based on manufacturing and nonclinical data. In 2025 CBER's leadership restricted the indications and required new postmarketing clinical studies. Pandemic influenza sits between the two: complete CMC and nonclinical data before approval, with clinical data possibly collected after approval.

Key takeaways

Legal mechanism. Under 21 CFR 601.12, a change with substantial potential to adversely affect identity, strength, quality, purity or potency needs a prior approval supplement before distribution. Examples include changes in formulation, source material, or a new master cell bank or seed ¹⁰¹¹⁰². The regulation does not mention "strain change" by name ¹⁰⁶. An approved comparability protocol can justify a reduced reporting category for a specified change ¹⁰⁷.
Seasonal influenza. After strain selection, manufacturers submit license supplements that include seed information and updated labeling. Production typically starts in parallel with supplement review; formulated bulks go to CBER for release, and vaccine is distributed only after supplement approval and lot release ⁵⁴⁹⁵⁵⁰. FDA has said new clinical data are not required for these updates because antigenic data are predictive and manufacturing experience is extensive ³⁷⁴³⁷⁵.
COVID-19, 2022. FDA said the influenza model could not yet be applied. It required manufacturer-specific clinical safety and immunogenicity data for modified vaccines ³⁷²³⁷³³⁷⁴.
COVID-19, 2023 to 2024. FDA based the 2023-2024 (XBB.1.5) and 2024-2025 (KP.2/JN.1) formula actions on manufacturing data and nonclinical immune-response data ⁵¹¹⁵¹²¹³⁶¹³⁹.
COVID-19, 2025 onward. CBER limited the indications to adults 65 and older and to younger people at high risk, with age floors and prior-vaccination conditions that differ by product. It also obtained postmarketing commitments for formula-specific immunogenicity studies and randomized trials in healthy 50 to 64 year olds ⁴⁶⁹⁴⁷¹⁴⁷²⁴⁷⁴⁴³⁸⁴⁴⁰.

The regulatory mechanism: supplements under 21 CFR 601.12

21 CFR 601.12 requires the applicant to assess the effects of any change before distributing product made with it. The applicant must show, through appropriate studies or validation, that the change has no adverse effect ¹⁰¹. It sets three reporting tiers:

Tier	Trigger	Timing
Prior approval supplement	Substantial potential for adverse effect (e.g., formulation, source material or cell line, new master cell bank or seed)	FDA approval before distribution ¹⁰²
CBE-30 supplement	Moderate potential	Submitted at least 30 days before distribution ¹⁰³¹⁰⁴
Annual report	Minimal potential	Within 60 days of the approval anniversary ¹⁰⁵

Every vaccine strain update covered below was handled as a prior-approval supplement to an existing BLA; MFLUSIVA's annual updates are filed under an approved comparability protocol. Each was approved before distribution.

Seasonal influenza: the established strain-change model

Strain selection

FDA's decision process aims to pick antigens likely to protect against the range of viruses that may co-circulate. The review covers:

surveillance and epidemiology;
genetic subclades and antigenic drift;
integrated laboratory analyses and human serum panels;
human serology and vaccine effectiveness;
whether candidate vaccine viruses are available, and their characteristics, for each manufacturing platform ²³⁴²³⁵²³⁶²³⁷²³⁸.

WHO issues Northern and Southern Hemisphere recommendations twice a year. VRBPAC then considers each component for U.S. vaccines ²⁴¹²⁴². Recent outcomes:

March 2023: H1N1, H3N2 and B/Victoria each passed 13-0-0. Keeping B/Yamagata passed 7 yes, 2 no, 4 abstain, mainly because it was uncertain whether the lineage had died out. No detections had been confirmed after March 2020 ²⁴³²⁴⁴.
October 2023: VRBPAC was asked whether to exclude B/Yamagata as soon as possible ²⁴⁵.
March 2024: FDA reported that manufacturers had filed trivalent regulatory materials, that approvals were on track for 2024-25, and that only trivalent vaccine would be distributed in the U.S. ²⁴⁶²⁴⁷.
March 12, 2026 (2026-2027 season): VRBPAC voted 7-0-0 on separate strain sets for egg-based vaccines and for cell- and recombinant-based vaccines. For example, the H3N2 virus was A/Darwin/1454/2025-like for egg-based vaccines and A/Darwin/1415/2025-like for cell and recombinant platforms ⁵⁶⁵.

Implementation and data package

FDA has described the sequence to VRBPAC as follows:

The committee votes on U.S. strains.
Manufacturers submit license supplements that add the recommended composition.
Updated vaccine is distributed after FDA approval ⁵⁴⁹.

The supplement includes seed information and updated labeling. Production of the new strains typically starts after strain selection, in parallel with supplement review. Formulated bulks go to CBER for release, and vaccine is distributed only after the supplement is approved and lots are released ⁵⁵⁰. FDA calls strain-change supplements to existing BLAs an established regulatory pathway ⁵⁵¹. Northern Hemisphere recommendations usually come in February or March, and supplement approvals usually in June or July ⁵⁵².

On clinical data, FDA's April 2022 VRBPAC briefing said updated seasonal influenza vaccines do not need new clinical data. FDA attributed this to predictive antigenic data and extensive manufacturing experience ³⁷⁴³⁷⁵. The 2007 guidance Clinical Data Needed to Support the Licensure of Seasonal Inactivated Influenza Vaccines covers clinical development for new seasonal vaccines, not annual updates ¹²²¹²³¹²⁴¹²⁵. It also states that it does not address chemistry, manufacturing and controls (CMC) ¹²⁶.

The trivalent transition

Removing B/Yamagata was a formulation change, so it was also handled by supplement:

Flucelvax (Seqirus), STN 125408/575, approved March 4, 2024: added a trivalent formulation for individuals 6 months and older ²⁸⁹⁴⁰¹.
Flublok: CBER review documents refer to STN 125285/580 as the "transition to trivalent formulations." The Flublok labeling review stated that quadrivalent study data were relevant to the trivalent product because the two share manufacturing processes and overlapping compositions ³³⁰³³¹.

mRNA seasonal influenza: a protocol-driven strain-change model

The MFLUSIVA (mRNA-1010, ModernaTX) BLA shows how CBER has set up annual strain changes for a new platform ⁵³⁸. The CMC review describes an approved comparability-protocol approach ⁵³²:

The RNA and LNP processes are verified for the updated strains.
Analytical comparability is assessed, and a representative drug-product lot is made combining updated and unchanged strains.
Processes stay the same apart from strain-specific starting materials.
Registered release, extended-characterization and stability tests and their acceptance criteria are not changed.
Mouse immunogenicity studies confirm a response to the updated strains.

Data generated under the protocol must be submitted in a prior approval supplement, and product made with the change can be distributed once that supplement is approved ³³⁵⁵³³. Lot release remains in place: final-container samples and protocols go to CBER, and lots cannot be distributed until CBER releases them ⁵³⁴. The protocol calls for nonclinical mouse data, not a new clinical study for each annual change ⁵³⁵³³⁵. The approval letter row is dated August 6, 2026 ³²². Before that, VRBPAC voted 9-0 on June 18, 2026 that benefits outweighed risks for adults 50 to 64 (traditional approval) and 65 and older (accelerated approval) ⁶³⁷. During review, FDA questioned whether quadrivalent immunogenicity data from Study P303 Part C could support the trivalent formulation, specifically whether the extra B strain caused enhancement or interference ⁵³⁶.

Pandemic influenza: CMC and nonclinical data before approval, clinical data possibly after

FDA has set out two pathways for already-licensed pandemic prototype vaccines (for example H5):

Scenario	Pre-approval data	Post-approval data
Update between pandemics (proposed process)	CMC data for quality and consistency, plus clinical immunogenicity and safety data ⁵⁰¹	Not specified
Strain change during a declared pandemic	Complete CMC information, plus nonclinical data supporting effectiveness ⁵⁰²⁵⁰³	FDA may require clinical immunogenicity and safety data after approval; manufacturers should plan to collect safety and effectiveness data during deployment ³⁷⁶⁵⁰³

Both pathways assume a licensed prototype vaccine for that subtype. The original licensure was supported by clinical safety and immunogenicity data, with effectiveness inferred from an efficacious licensed seasonal vaccine made by the same process. A vaccine against a new subtype would not be treated as a simple strain change ⁵⁰³⁵⁰⁴. The pandemic influenza guidance gives numerical immunogenicity criteria for comparing a new vaccine with a comparator: the upper bound of the two-sided 95% CI on the GMT ratio (comparator/new) should not exceed 1.5, and the upper bound on the seroconversion-rate difference (comparator minus new) should not exceed 10 percentage points ²⁸¹.

A recent example is AREPANRIX (ID Biomedical Corporation of Quebec), STN 125419/166, approved April 28, 2026. It changed the strain from A/Indonesia/05/2005 (H5N1) to A/Astrakhan/3212/2020-like (H5N8, clade 2.3.4.4b), with labeling revisions ⁶³²⁹. The approval letter also describes changes to the proper name and indication ³²⁹.

COVID-19 vaccines: from clinical immunobridging to the influenza-like model, then a reset

2022: clinical data required

In 2022 FDA told VRBPAC that limited experience meant it required manufacturer-specific clinical safety and immunogenicity data to authorize a modified COVID-19 vaccine. Evidence of effectiveness would rely on neutralizing antibody responses to relevant variants plus the demonstrated effectiveness of the original vaccine ³⁷²³⁷³. The April 2022 briefing said manufacturing information alone was not enough. FDA asked the committee what further experience might eventually allow a manufacturing-only approach ³⁷⁴³⁷⁵.

FDA's summary of the BA.1-containing mRNA candidates was as follows:

Monovalent and bivalent candidates gave statistically superior BA.1 neutralizing GMTs compared with the prototype vaccines, and ancestral-strain responses did not appear to drop ⁴⁵⁹.
For the Pfizer-BioNTech candidates, BA.1 GMT ratios against the 30 µg prototype were 2.23 and 3.15 for the 30 and 60 µg Omicron monovalent candidates, and 1.56 and 1.97 for the bivalent candidates ⁴⁶¹.
FDA said the human data were limited to neutralizing antibodies, mostly after a fourth dose, and that some assays had not finished validation ⁴⁶⁰.

On August 31, 2022, FDA authorized BA.4/BA.5 bivalent boosters based on the totality of evidence ⁵⁰⁹¹⁴⁰⁵¹⁰:

safety and effectiveness data for the original monovalent vaccines;
clinical safety and immune-response data for the investigational BA.1 bivalent vaccines, which FDA considered relevant to the BA.4/BA.5 formulations;
nonclinical data for the BA.4/BA.5 bivalent vaccines.

2023 to 2024: manufacturing and nonclinical data

2023-2024 formula (XBB.1.5). VRBPAC recommended a monovalent XBB-lineage update and agreed on XBB.1.5 ¹⁷⁶. On September 11, 2023, FDA approved the updated vaccines for ages 12 and older and authorized them for 6 months through 11 years ⁵¹³. FDA based these actions on manufacturing data for the formula change and nonclinical immune-response data, plus experience with earlier mRNA vaccines. It also cited data showing the updated vaccines neutralized EG.5 and BA.2.86 about as well as earlier vaccines neutralized their matched variants ⁵¹¹⁵¹². The supplements were Comirnaty STN 125742/276 ⁸⁹ and Spikevax STN 125752/68, both approved September 11, 2023 ²³⁴.
2024-2025 formula (JN.1 lineage). VRBPAC voted unanimously for a monovalent JN.1-lineage composition. FDA preferred KP.2 where feasible ¹⁷⁷¹⁷⁸. On August 22, 2024, FDA approved updated Comirnaty and Spikevax (KP.2) for ages 12 and older and authorized them for 6 months through 11 years ¹³⁴¹³⁵. FDA said it assessed manufacturing and nonclinical data and that the manufacturing process was similar to earlier formulas ¹³⁶. The Novavax 2024-2025 (JN.1) vaccine received an EUA on August 30, 2024, on the same stated basis ¹³⁷¹³⁸¹³⁹.

The inputs FDA put to VRBPAC for strain selection were:

manufacturer postvaccination human serology against newer variants;
manufacturer nonclinical immunogenicity studies of candidate updated spike antigens;
virus surveillance and antigenic characterization;
vaccine-effectiveness and clinical-immunogenicity studies ¹⁸¹¹⁸².

FDA also cited manufacturing lead times. Changing an mRNA vaccine generally takes less time than changing a protein-subunit vaccine ¹⁸³.

The Nuvaxovid BLA: how immunobridging was framed

The Nuvaxovid review shows the quantitative immunobridging criteria CBER applied:

Study 313. A single dose of monovalent XBB.1.5 vaccine in vaccine-naive participants was compared with an XBB.1.5 booster in previously mRNA-vaccinated participants. The endpoint was neutralizing antibody to XBB.1.5 at Day 28 ³⁹²³⁹³.
- Noninferiority criteria: lower 95% CI bound above 0.67 for the GMT ratio and above -10 percentage points for the seroresponse rate (SRR) difference ³⁹⁴.
- Results: GMT ratio 1.8 (95% CI 1.4 to 2.2); SRR 74.3% vs 64.3%, a difference of 10.0 points (95% CI 2.6 to 17.4). Both criteria were met ³⁹⁵³⁹⁶³⁹⁷³⁹⁴.
Study 311. For an earlier BA.5-containing bivalent formulation, the plan required superiority against BA.5 (lower bound above 1.0) and noninferiority against the ancestral strain (lower bound above 0.67) ³⁹⁹.
Recommendation. Reviewers recommended approval based on the totality of evidence:
- original-formula efficacy;
- immune-response bridging for strain updates;
- Study 313;
- nonclinical data showing higher neutralization of JN.1-related sublineages by the JN.1 formula than by the XBB.1.5 formula ⁷²⁷³⁷⁴⁷⁵⁷⁶.

The clinical review inferred effectiveness of the updated formula from immunogenicity data. It treated data on earlier formulations as relevant because the manufacturing process was similar and the antigen and adjuvant doses were the same. It also noted that direct effectiveness data against Omicron-sublineage outcomes were lacking ⁴¹³.

Nuvaxovid received BLA approval on May 16, 2025, for adults 65 and older and people 12 to 64 with at least one high-risk condition ³¹⁰³¹¹. The postmarketing commitments include:

a randomized controlled efficacy trial in 50 to 64 year olds ³¹²;
a real-world effectiveness study of the 2024-2025 formula and anticipated later strain updates ³¹³.

The 2025-2026 formula sBLA was approved August 27, 2025 ⁷⁷⁷⁸.

2025: restricted indications and new postmarketing studies

VRBPAC voted 9-0-0 for a monovalent JN.1-lineage composition for 2025-2026 ¹⁷⁹¹⁸⁰. The approvals that followed changed both the indication and the post-approval evidence expectations.

Comirnaty. The Center Director Decisional Memo, dated August 25, 2025, found that efficacy assessments of updated formulas had relied on immunogenicity and observational data rather than new efficacy trials. It said antibody increases do not by themselves establish clinical benefit, and observational estimates can be confounded ⁴⁶⁹⁴⁷⁰⁴¹⁶. CBER approved the vaccine for adults 65 and older and for people 5 to 64 with a high-risk condition ⁴⁷¹. Supplement STN 125742/656 was approved August 27, 2025 ⁸⁹. Pfizer agreed to three postmarketing commitments:

a safety and immunogenicity study of the 2025-2026 formula in approved adult and adolescent risk groups, powered for prespecified immunogenicity testing in each age group ⁴⁷²⁴⁷³;
a randomized, double-blind, saline-placebo-controlled efficacy and safety trial in 50 to 64 year olds without high-risk conditions ⁴⁷⁴⁴⁷⁵;
a safety and immunogenicity study in high-risk children 5 to 11 ⁴⁷⁶.

mNEXSPIKE. The Center Director Override Memo (May 30, 2025) limited the indication to previously vaccinated people who are 65 or older, or 12 to 64 with a high-risk condition ⁴³⁸⁴³⁹. The pivotal trial showed noninferiority to an earlier Spikevax formulation whose efficacy against later subvariants had not been established in randomized trials ⁴⁴⁰. A randomized efficacy commitment in 50 to 64 year olds was agreed ⁴³⁹⁴⁴¹. The 2025-2026 formula supplement (STN 125835/6) was approved August 27, 2025, with further safety, immunogenicity and safety-study commitments ⁴⁴²⁴⁴³⁴⁴⁴.

Spikevax. STN 125752/305 (2025-2026 formula) was approved August 27, 2025 ²³⁴⁴⁵. The agreed commitments were a safety and immunogenicity study in older and higher-risk participants, and a randomized, placebo-controlled safety study in 50 to 64 year olds without high-risk conditions ⁴⁴⁹. FDA later released the August 2025 randomized safety-study commitments for both Moderna products because of operational and feasibility problems. It said new commitments would replace them; the release is not a finding that the studies were completed ⁴⁵⁰⁴⁵¹.

2026-2027 formula

On May 28, 2026, VRBPAC voted 8 yes, 0 no, 1 abstain for the JN.1-lineage XFG variant as the preferred strain for a monovalent 2026-2027 vaccine ⁶³⁹⁶⁴⁰⁶⁴¹. FDA's briefing again listed the main inputs as postvaccination human serology and nonclinical immunogenicity of updated spike candidates ⁶⁴²⁶⁴³. On August 27, 2026, CBER approved 2026-2027 formula supplements for:

Spikevax (BL 125752/381) ⁵⁹⁷⁵⁹⁸;
Nuvaxovid (BL 125817/131) ⁵⁹⁹;
mNEXSPIKE (BL 125835/166) ⁶⁰⁰;
Comirnaty (BL 125742/860) ⁶⁰¹.

The approval records retrieved for these supplements do not state the data basis ⁵⁹⁰⁵⁹¹⁵⁹²⁵⁹⁴.

Side-by-side comparison

Dimension	Seasonal influenza	Pandemic influenza (licensed prototype)	COVID-19 (current practice)
Strain selection	WHO recommendation, then VRBPAC vote by component and platform ²⁴¹²⁴²⁵⁶⁵	Not addressed in retrieved sources	VRBPAC vote on lineage and preferred variant ¹⁷⁹⁶³⁹
Submission	License supplement with seed information and labeling ⁵⁵⁰	Pre-approval supplement ⁵⁰²	sBLA "to include the [year] Formula" ⁸²
CMC	Supplement, then CBER lot release ⁵⁵⁰; protocol-based analytical comparability for mRNA ⁵³²	Complete CMC information ⁵⁰²	Manufacturing data for the formula change ⁵¹¹¹³⁶
Nonclinical	Not required for established egg, cell and recombinant strain changes; mouse immunogenicity is specified only in the MFLUSIVA comparability protocol ³⁷⁴⁵³²	Nonclinical effectiveness data before approval ⁵⁰²	Nonclinical immune-response data ⁵¹¹¹³⁶
Clinical before approval	Not required ³⁷⁴³⁷⁵	Not required during a declared pandemic; required for updates between pandemics ⁵⁰¹⁵⁰²	Not required for 2023 to 2024 updates ⁵¹¹¹³⁶; 2022 relied on BA.1 clinical data ⁵⁰⁹
Post-approval clinical evidence	Not addressed in retrieved sources	May be required ³⁷⁶⁵⁰³	Formula-specific immunogenicity studies and randomized trials in healthy 50 to 64 year olds ⁴⁷²⁴⁷⁴⁴⁴⁹

Practical implications for sponsors

Plan the platform-level justification early. FDA's case for reduced data in each setting rests on manufacturing sameness: same process, same antigen and adjuvant doses, unchanged release specifications ⁴¹³⁵³²¹³⁶. A comparability protocol approved at licensure, like MFLUSIVA's, sets in advance what each annual prior approval supplement will contain ⁵³²³³⁵.
Expect manufacturing data and nonclinical immunogenicity to carry routine COVID-19 formula updates before approval, as in 2023 to 2026. Human serology and nonclinical cross-neutralization data also drive strain selection ¹⁸¹⁶⁴².
For COVID-19, budget for postmarketing clinical work. Since 2025, formula approvals have come with commitments for formula-specific immunogenicity studies in the indicated populations and randomized studies in healthy middle-aged adults ⁴⁷²⁴⁷⁴⁴⁴⁹⁴⁴³.
Use established immunobridging criteria when clinical bridging is needed. The Nuvaxovid program used a noninferiority bound of 0.67 for the GMT ratio and -10 percentage points for the SRR difference ³⁹⁴³⁹⁹. The pandemic influenza guidance uses different margins: 1.5 for the GMT ratio and 10 percentage points for the seroconversion difference, each expressed as comparator minus or over new vaccine ²⁸¹.

Open questions

The current FDA guidance Development and Licensure of Vaccines to Prevent COVID-19 (October 2023) does not, in the passages retrieved, set out a variant- or strain-change-specific data package ³⁷¹³. In particular, it gives no numerical GMT-ratio or seroresponse margins for modified vaccines ³⁹⁴⁰. CBER's approval records for the 2025-2026 and 2026-2027 COVID-19 formula supplements also do not detail the strain-specific nonclinical and CMC data reviewed ⁷¹⁵⁹⁰. Readers who need that level of detail should look at the individual review memoranda for each supplement.