A multicancer blood test did not reduce late-stage cancer after three screening rounds

Design

Randomised controlled screening trial, initially blinded

Population

142,250 adults aged 50–77 in England

Primary outcome

Incidence of stage III or IV cancer across 12 prespecified types

Effect

IRR 1.03 (95% CI 0.92–1.14); stage IV IRR 0.86 (0.74–1.00)

The NHS-Galleri trial, published in NEJM on 22 September 2026, took blood from 142,250 people aged 50 to 77 in England at up to three annual visits. Half had their blood tested with a multicancer early-detection assay alongside usual care; the other half had samples stored untested.

The primary endpoint, the rate of stage III or IV disease across 12 prespecified cancers, did not differ (incidence rate ratio 1.03). Stage IV cancers alone were somewhat lower in the tested group (ratio 0.86), with a confidence interval reaching 1.00. Adverse events from the trial were rare and none were serious.

A screening test earns its place by moving diagnoses to an earlier, curable stage. On its main measure, this one did not. The stage IV signal is a secondary endpoint and could strengthen or fade with time. For now, multicancer blood tests sold directly to patients are not a substitute for established cervical, breast and colorectal screening, and a negative result should not falsely reassure.

• Advise patients that multicancer blood tests have not been shown to reduce late-stage cancer.
• Keep patients on established cervical, breast and colorectal screening regardless of a multicancer test result.
• Do not let a negative multicancer blood test delay investigation of new symptoms.
• Expect longer follow-up from this trial before any change to screening policy.

Why it matters

Patients are already being sold these tests, and this is the evidence to discuss with them.

Don't overread it

The stage IV reduction is a secondary endpoint at the edge of significance, not a demonstrated benefit.

The statistics, in plain English

An incidence rate ratio of 1.03 means late-stage cancers were as common in the tested group as in the untested group. The stage IV ratio of 0.86 has an interval touching 1.00, so it could be chance, and it was not the main question the trial was designed to answer.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.