A United States advisory panel has recommended approval of a multi-cancer early detection blood test. The Molecular and Clinical Genetics Panel of the Medical Devices Advisory Committee met on 23 September 2026. It reviewed GRAIL’s Galleri test. The assay reads methylation patterns in cell-free DNA. The panel agreed the test is safe. It divided on whether the test works.
All ten voting members backed safety. Only six of the ten backed effectiveness, with four against. Seven voted that the benefits outweigh the risks. Two voted against and one abstained. The distance between those three votes is the story. It also maps onto an evidence argument England has already had.
What the FDA panel was asked to decide
The panel answered three questions, one each on safety, effectiveness and benefit-risk. Each was framed around patients who meet the proposed indication for use.
That indication is precise. The FDA executive summary describes a test for “screening for the early detection of multiple types of cancer”. The eligible population is “adults aged 50 years or older”. A positive result also returns a Cancer Signal Origin prediction, naming the likely site. The test is positioned as an addition to guideline-recommended screening, never a replacement.
GRAIL submitted its premarket approval application on 29 January 2026. The agency is not bound by the panel recommendation. GRAIL’s statement on the vote says a final decision is expected in the coming months.
Why the effectiveness vote on multi-cancer early detection split
The agency set the agenda itself. Its briefing document asked the panel to weigh one issue in particular. That issue was the “clinical significance of the inter-study variability” in the test’s performance estimates. Two pivotal studies supported the application, and they did not agree.
PATHFINDER 2 ran in North America, and LSDN covered its interim findings in October 2025. Sensitivity at twelve months was 35.0 per cent. The 95 per cent confidence interval ran from 29.8 to 40.5. NHS-Galleri ran in England. Its sensitivity was 31.6 per cent, on an interval of 28.5 to 34.8.
By cancer type the gap widened considerably. Head and neck sensitivity was 81.2 per cent in PATHFINDER 2. The same figure was 42.9 per cent in NHS-Galleri. The agency asked whether any cancer types present new risks because of that uncertainty.
The FDA also questioned the phrase at the centre of the product. It asked whether the stage data support a claim of early detection. The alternative is a broader claim of cancer detection. The stage figures explain the question. In NHS-Galleri the test detected 20.5 per cent of stage I and II cancers. In PATHFINDER 2 it detected 24.9 per cent.
The harms side of the multi-cancer early detection ledger
Specificity was high in both studies. PATHFINDER 2 recorded 99.85 per cent and NHS-Galleri 99.74 per cent. False positive rates were 0.15 and 0.26 per cent. The number needed to screen to find one cancer was 202 and 196 respectively.
Those figures look reassuring until they are converted into procedures. Of 218 PATHFINDER 2 participants with a positive result, 90 had no cancer diagnosis within twelve months. Of those 90, 43 (47.8 per cent) still underwent at least one invasive procedure. Nine participants had adverse events during diagnostic evaluation. The FDA records none of the study-related events as serious.
The agency flagged the opposite failure mode as well. It warned that a negative result could produce false reassurance. It also warned that some people might forgo guideline-recommended screening. The stated reason was the convenience of a blood draw.
Much of the argument for multi-cancer early detection rests on cancers with no screening programme. Even where surveillance exists, adherence gaps remain. A contributor analysis on LSDN examined surveillance gaps in liver cancer detection. The FDA briefing adds a figure that cuts both ways. For cancer types that already have guideline screening, PATHFINDER 2 sensitivity was 17.7 per cent in screening-eligible people. It was 31.6 per cent in those not eligible. NHS-Galleri showed the reverse, at 28.8 and 24.8 per cent.
England ran a 142,000-person randomised trial of multi-cancer early detection
NHS-Galleri recruited 142,250 people aged 50 to 77 in England. Participants were randomised to the test or to standard NHS care. Blood was collected across three annual screening rounds. The FDA analysis covers 70,323 intervention-arm participants with valid results.
The primary endpoint was a reduction in combined stage III and stage IV cancer incidence. It was not met. Queen Mary University of London stated that result plainly in its announcement of 23 September 2026. The university ran the trial through the Cancer Research UK Cancer Prevention Trials Unit.
Several secondary findings pointed the other way. There were fewer stage IV cancers in the intervention group. There were more stage I and stage II diagnoses. There were also fewer cancers diagnosed after an accident and emergency presentation.
Around one in 100 participants tested positive in each round. Across the three rounds, 937 of them were subsequently diagnosed with cancer. Positive predictive value ranged from 46 to 58 per cent across the rounds.
Professor Peter Sasieni set the two findings side by side. He is Professor of Cancer Epidemiology at Queen Mary University of London. He was also one of three scientific leads on the trial. “Although we did not see a reduction in our primary endpoint,” he said, “we observed fewer stage IV cancers.” He paired that with “an increase in cancers diagnosed at earlier stages”. He also said that “a positive test cannot be ignored”, because about half of positives proved to be cancer.
The timing was tight. The NHS-Galleri primary result appeared in the New England Journal of Medicine on 22 September 2026. Performance papers from both trials appeared in Nature Medicine the same day. The panel met the following day.
The evidence bar England set for multi-cancer early detection
The UK National Screening Committee published a position statement in The BMJ on 25 May 2026. It sets out what the committee expects from multicancer detection tests. That standard applies before any recommendation for a national programme.
The statement is direct about study design. It calls for “adequately powered and well designed randomised trials”. Those trials should compare “those offered screening and those not offered screening”. The committee describes that route as the most reliable way to test whether screening reduces mortality.
It then names the surrogate it will accept. The committee points to “reductions in the incidence of late-stage (stage III and IV) cancer”. It calls that “the most promising surrogate outcome for cancer specific mortality”. It adds a caution about tumour biology. Tests may “preferentially detect cancers with more aggressive biology”, which would weaken that link.
That surrogate is the endpoint NHS-Galleri did not meet. LSDN makes that observation as a reading of two published documents. It is not a UK National Screening Committee position on the Galleri test. The committee has published none.
The route into the NHS runs through the same body. The National Cancer Plan for England was published in February 2026. It ties action to evidence “as judged by UK NSC”. The threshold is evidence that “shows that MCEDs are effective”. It adds a second condition, “as the fiscal position permits”. The plan then promises “a fully worked up implementation plan” delivered “at scale through phlebotomy services”. Two conditions sit in that commitment, and only one of them is scientific.
LSDN ran two searches on 25 September 2026. The UK NSC recommendations service listed no recommendation on multicancer detection. The committee blog carried no 2026 post on multi-cancer early detection or on the NHS-Galleri trial.
Commercial position and market reaction
Galleri already sells in the United States. GRAIL’s own patient pages give a list price of 949 US dollars. They also state that the test is prescription only. The same pages say it “is not covered by Medicare or most health insurance plans at this time”.
Those pages carry a regulatory disclosure too. They state that the test “has not been cleared or approved by the Food and Drug Administration”. A premarket approval would change that status. It would also change the reimbursement argument for multi-cancer early detection.
Investors moved before the panel did. STAT reported that GRAIL shares surged 33 per cent on 21 September 2026. That followed the FDA posting its meeting documents. Shares rose again after the vote. Outlets did not agree a figure, so LSDN reports direction only.
What the panel asked GRAIL to change
The recommendation came with conditions on language. NPR reported that the panel advised a clear instruction in marketing materials. People should be told to continue routine cancer screening. NPR also reported advice against describing Galleri as an early detection tool. That is the phrase sitting in the proposed indication.
Not every voice at the meeting was supportive. NPR reported comments from Phil Castle. He directs the National Cancer Institute’s Division of Cancer Prevention. He said that “routine approval and broad population rollout would risk harm and waste resources”. LSDN carries that comment as reported by a single outlet.
What to watch next
Three things now decide how far multi-cancer early detection travels. The first is the FDA decision, and the final wording of the indication. The second is whether the agency attaches a post-approval study. Its briefing document describes none.
The third is British. England holds a 142,000-person randomised dataset. It has also written its adoption pathway down. Whether that pathway opens depends on a committee that has not yet spoken. It depends on a fiscal test as well as a scientific one.














