The NHS-Galleri blood test found cancers but did not reduce late-stage diagnoses
Finding a cancer and changing its outcome are not the same thing. The largest trial of a multicancer blood test shows why.
NHS-Galleri was a randomised trial in England. It tested whether adding a multicancer blood test to usual care would cut late-stage cancer diagnoses. It enrolled 142,250 people aged 50 to 77 with no symptoms of cancer. Each gave blood at up to three yearly visits.¹
Half had their blood tested (71,122). The other half had their blood stored but not tested (71,128). A positive test led to referral for NHS tests. The test's guess at where the cancer started guided the referral. Both groups carried on with usual care.¹ ²
The main outcome was the rate of stage III or IV cancer across 12 cancer types chosen in advance. These were lung, head and neck, bowel, pancreas, myeloma, liver and bile duct, stomach, oesophagus, anus, lymphoma, ovary and bladder. Stage IV cancer in the same 12 types was the key secondary outcome.¹
The results
After three rounds of screening:¹
Stage III or IV cancer: no significant difference between the groups. Incidence rate ratio (IRR) 1.03 (95% CI 0.92 to 1.14, p=0.63).
Stage IV cancer: IRR 0.86 (95% CI 0.74 to 1.00). The authors do not call this significant.
Safety: fewer than 1% of people had side effects linked to the trial. None were serious.
How well the test performed
A companion paper reports results for each round:²
Round 1: 722 positive results from 70,325 tests (1.03%). Positive predictive value (PPV) 58.0%.
Round 2: 518 from 64,498 (0.80%). PPV 50.4%.
Round 3: 561 from 62,323 (0.90%). PPV 45.8%.
So the chance that a positive result meant cancer fell with each round. The test picked up 27% to 37% of all cancers diagnosed within a year of a test. In round 1, about two-thirds of the missed cancers were stage I or II.² The negative predictive value was about 99% in every round. Even so, about 1 in 100 people with a negative result were diagnosed with cancer in the year that followed.²
The authors conclude that the test did not cut stage III or IV cancer after three rounds. They say longer follow-up is needed for both the main result and the stage IV result.¹ GRAIL, which makes the test, funded the trial. It helped design the study and draft the companion paper.²
What does this mean in primary care?
The test found cancers. But it did not cut late-stage diagnoses, the outcome the trial was designed to test.
The main outcome was late-stage diagnosis, not cancer deaths.
A positive result led to a full work-up. The chance it meant cancer fell from 58% in round 1 to 46% in round 3.
A negative result did not rule cancer out.
Patients who are thinking about paying for the test may ask about these results. The trial shows that finding cancer and improving outcomes are separate questions.
Intended for UK healthcare professionals only. This article reports on published clinical data and does not constitute medical advice.
References
Sasieni P, Johnson P, Round T, et al. Effect of screening with multicancer early-detection test on late-stage cancer diagnosis. N Engl J Med. 2026;395:1374–1386. doi:10.1056/NEJMoa2505723
Neal RD, Dolly S, Johnson P, et al. Performance of a multi-cancer early detection test in the randomized controlled NHS-Galleri trial. Nat Med. 2026;32:3770–3780. doi:10.1038/s41591-026-04652-8