mRNA vaccines after COVID: what the flu, RSV and cancer trials measured

Three questions about mRNA vaccines after COVID sound like one: is the mRNA flu shot better than the shot you already get, does one RSV dose cover more than a single winter, and can a "cancer vaccine" do anything about a tumour that already exists. Each has a published trial behind it, and each answers something narrower than its headline. Here is what was measured, in whom, and for how long.

This is an information article, not medical advice. It does not replace a consultation with a doctor, and nothing here is a recommendation for or against any vaccination.

Does the mRNA flu shot beat the one I already get?

The pivotal trial was built not against placebo but against a licensed standard-dose flu vaccine, which makes the result a relative figure.

Moderna's mRNA-1010 was tested in a phase 3, double-blind, active-controlled trial in adults 50 and older: 40,703 people vaccinated (20,350 mRNA, 20,353 comparator), median follow-up 181 days. The primary endpoint was relative vaccine efficacy against RT-PCR-confirmed, protocol-defined influenza-like illness, from at least 14 days after vaccination to the end of the 2024–25 northern season. Cases occurred in 411 of 20,179 (2.0%) versus 557 of 20,124 (2.8%) — relative efficacy 26.6% (95% CI 16.7–35.4), meeting the trial's non-inferiority and superiority criteria. By strain the point estimates ran from 22.2% for A/H3N2 to 29.6% for A/H1N1, and the interval for B/Victoria crossed zero; in adults 65 and older the estimate was 27.4% (12.1–40.0). Solicited reactions — injection-site pain, fatigue, headache — were more common in the mRNA group, per secondary accounts of the same trial.

A separate modified-mRNA flu vaccine, tested against a control vaccine in 18,476 adults, showed relative efficacy of 34.5% (95% CI 7.4–53.9) from 57 versus 87 cases; no approval for it was verified in the records checked.

The key to both numbers: "26.6% relative efficacy" means fewer laboratory-confirmed influenza illnesses than among people who got the standard-dose comparator, on top of whatever that comparator already did. It is not 27% versus no vaccination, and it says nothing about hospital admission or death.

A US approval of mRNA-1010 (mFlusiva) for adults 50 and older was reported on 5 August 2026, with the 65-and-older clearance called accelerated and tied to a required post-marketing study — trade press, not an approval letter, so treat it as reported.

Then what was the "100% protection" challenge study?

A different design entirely. In a phase 2a, double-blind challenge trial, healthy adults aged 18–55 received the modified-mRNA vaccine or a standard quadrivalent vaccine and, 30 days later, were deliberately exposed to one influenza A/H1N1 strain. Per protocol — 55 on mRNA, 48 on the standard vaccine, 52 unvaccinated controls from a separately challenged cohort — symptomatic influenza occurred in 0%, 4.2% and 26.9%; efficacy against the unvaccinated controls was 100% (95% CI 75.2–100) and 84.5% (43.4–96.0). There were no serious adverse events.

The honest part of that "100%" is the interval, which starts at 75.2%: the groups are small and few controls fell ill. One strain, one deliberate exposure, about fifty healthy adults per group. Such a study shows an immune response can block a known virus under controlled conditions — not what happens across a flu season.

Is there a universal flu vaccine yet?

Not in human efficacy data, on the records checked. The most advanced mRNA work aimed at the conserved part of the virus — the hemagglutinin "stalk", which varies far less between strains than the head — is in monkeys. Rhesus macaques that already had influenza immunity were given nucleoside-modified mRNA encoding chimeric hemagglutinins: they produced stalk-binding antibodies, their serum protected mice in a lethal challenge with a different strain, and bone-marrow plasma cells remained detectable at 8 months.

One trap, because the two get merged online: a completed phase 1/2a trial in 170 people, often cited as a "universal flu vaccine" study, tested a protein stem vaccine, not mRNA.

How long does one RSV dose hold?

mRNA-1345 was tested in a phase 2/3, randomised, observer-blind, placebo-controlled trial in adults 60 and older: 17,793 vaccine versus 17,748 placebo, one dose. The primary endpoints were a first episode of RSV-associated lower-respiratory disease with at least two signs or symptoms, and with at least three.

At the primary analysis, median follow-up 112 days, efficacy was 83.7% (95.88% CI 66.0–92.2) for the two-symptom endpoint — from 9 versus 55 cases — and 82.4% (96.36% CI 34.8–95.3) for three; serious adverse events occurred in 2.8% of each group.

The duration question has a measured answer. At a median 8.6 months, efficacy was 63.3% (48.7–73.7) for the two-symptom endpoint and 63.0% (37.3–78.2) for three — lower as follow-up lengthened. The product was approved in the United States on 31 May 2024 for adults 60 and older; a June 2025 extension to adults 18–59 at increased risk is reported, with no separate efficacy figure found.

Can a "cancer vaccine" treat a tumour that already exists?

Neither of the two best-known mRNA cancer programmes tested that: both were given after surgery, alongside other treatment.

In KEYNOTE-942, a randomised, open-label phase 2b trial, 157 people with completely resected stage III/IV melanoma — disease-free at entry — received an individualised mRNA vaccine encoding up to 34 neoantigens from their own tumour plus pembrolizumab (107), or pembrolizumab alone (50). Recurrence or death occurred in 24 of 107 (22%) versus 20 of 50 (40%): hazard ratio 0.561 (95% CI 0.309–1.017), two-sided p=0.053, with grade 3 or higher treatment-related adverse events in 25% versus 18%. That interval includes 1 — no difference — which makes the result a reason to run a phase 3 trial, not a treatment option. The product is not approved. A five-year update circulating in trade press is left out: it was not checked against the paper.

The pancreatic study is smaller: phase 1, single-arm, 16 people, all of whom had surgery, a checkpoint antibody, the vaccine and chemotherapy. At a median 3.2 years, the 8 who mounted a strong vaccine-induced T-cell response had median recurrence-free survival not reached, against 13.4 months in the 8 who did not (log-rank P=0.007); two responders recurred anyway. Responder status was determined after treatment, so this can mean the vaccine helped — or that people able to raise those T cells were already less likely to recur. The quoted "clones lasting decades" figure is a model projection.

What these results do not show

• Relative flu efficacy is measured against another vaccine, and says nothing about no vaccination, death or hospital admission. The comparator was a standard-dose product, not the high-dose or adjuvanted ones often used from age 65.

• RSV figures come from adults 60 and older, against medically defined lower-respiratory disease rather than every winter cold; the three-symptom interval ran from about 35% to 95% at the first analysis.

• The cancer results are adjuvant settings, in 157 and 16 people, added to treatment already working — nothing about metastatic disease, other cancers, or replacing existing therapy.

What to do with this

The practical move is to ask which product and which trial. "mRNA vaccine" now covers a licensed seasonal flu shot, a licensed RSV shot, an unapproved flu candidate and two investigational cancer programmes — four different evidence situations.

For a flu or RSV decision, the comparison that matters is with the specific product you would otherwise get, in your age group — a conversation for the clinician who knows your history. The RSV duration data is worth raising directly: measured efficacy was lower at 8.6 months than at about four, and what follows for revaccination timing is a clinical judgement.

If a cancer vaccine is the question, the honest framing is that the route to one today is a clinical trial, discussed with the treating oncologist.

Sources

• Leroux-Roels I et al. Efficacy and Safety of an mRNA Seasonal Influenza Vaccine in Adults. N Engl J Med. 2026;394:1803–1813. DOI 10.1056/NEJMoa2516491 · registry record NCT06602024

• Fitz-Patrick D et al. Efficacy, Immunogenicity, and Safety of Modified mRNA Influenza Vaccine. N Engl J Med. 2025;393:2001–2011. DOI 10.1056/NEJMoa2416779 · registry record NCT05540522

• Lindert KA et al. Human challenge trial of a nucleoside-modified mRNA influenza vaccine. NEJM Evid. 2025;4:EVIDoa2500087. PMID 41259791 · registry record ISRCTN13789612 (the registry's own page was redirecting to a holding page when this was checked; the record itself is retrievable through the WHO registry API)

• Styles TM et al. Chimeric hemagglutinin-based universal influenza mRNA vaccine in rhesus macaques. Cell Rep Med. 2025;6:102369. PMID 41005301

• Experimental universal influenza vaccine INFLUENZA G1 mHA, phase 1/2a — a protein stem vaccine, not mRNA. NCT05901636

• Wilson E et al. Efficacy and Safety of an mRNA-Based RSV PreF Vaccine in Older Adults. N Engl J Med. 2023;389:2233–2244. PMID 38091530 · registry record NCT05127434

• Ma C et al. Immune correlates analysis of the mRNA-1345 RSV vaccine efficacy trial. Nat Commun. 2025;16. DOI 10.1038/s41467-025-61153-x

• Weber JS et al. mRNA-4157 (V940) plus pembrolizumab versus pembrolizumab in resected melanoma (KEYNOTE-942): randomised phase 2b. Lancet. 2024;403:632–644. PMID 38246194

• Sethna Z et al. RNA neoantigen vaccines prime long-lived CD8+ T cells in pancreatic cancer. Nature. 2025;639:1042–1051. PMID 39972124