A blood test may detect a cancer signal before a person develops symptoms. Showing that a screening programme helps people live longer requires a different result. The NHS-Galleri primary-results abstract does not report a mortality benefit, and its main late-stage cancer endpoint was not met.
It does report findings worth examining, including more screen-detected cancers and fewer stage IV cancers. Keeping those outcomes separate is essential to understanding what the trial has shown so far.
What NHS-Galleri set out to test
The ASCO conference abstract describes 142,924 enrolled asymptomatic adults aged 50–79 in England. Participants were randomized to a group whose blood was tested using Galleri, or a control group whose blood was not tested, across up to three annual rounds.
Galleri looks for a shared cancer signal in circulating cell-free DNA. A signal leads to further diagnostic investigation; it is not itself a confirmed cancer diagnosis.
The primary endpoint concerned the incidence of stage III or IV cancers across 12 prespecified cancer types. This was a test of whether adding screening changed the occurrence of late-stage diagnoses, rather than simply whether the blood test found anything.
The main endpoint and the secondary finding
The primary endpoint was not met: the abstract reports 706 combined stage III/IV cancers in the intervention group and 688 in the control group. The incidence-rate ratio was 1.03, with a 95% confidence interval of 0.92–1.14.
Stage IV incidence across the same 12 prespecified cancer types was a separate, key secondary endpoint. There were 342 stage IV cancers in the intervention group and 397 in the control group, with an incidence-rate ratio of 0.86. Its reported confidence interval was 0.744–0.998.
These results were assessed approximately three years after the last participant was randomized. They answer different questions. A reduction in stage IV diagnoses can coexist with a primary endpoint that combines stages III and IV showing no reduction. One must not be substituted for the other.
The abstract alone also does not provide the full statistical analysis needed to establish how secondary findings should be interpreted after the primary endpoint was missed. The stage IV result should therefore remain identified as a secondary finding, rather than being promoted into an unqualified declaration of trial success.
More detection is not the final measure of benefit
The abstract reports more screen-detected cancers and changes in how cancers presented. These are meaningful descriptions of the diagnostic pathway. They do not, on their own, establish fewer deaths.
The US National Cancer Institute explains why multi-cancer screening needs trials that follow people over time. Potential benefits must be considered alongside false alarms, investigations and the detection of cancers that might not otherwise have caused harm. Finding disease earlier is not enough if the net outcome for screened people does not improve.
The distinction also matters when reading test-performance measures. The NHS-Galleri abstract reports a positive predictive value of 52%. A positive screening signal is therefore not synonymous with a cancer diagnosis, even in this study’s setting. That measure is different from the effect of offering screening to a population.
What would answer the headline question?
Evidence of a stage shift can support further investigation of the screening approach. Evidence of improved survival for people offered screening needs outcomes that actually address mortality, alongside the programme’s harms and follow-up burden.
The available NHS-Galleri source here is a conference abstract, with GRAIL identified as the research sponsor. It is not the full trial report or a report of mortality outcomes.
For now, the supported answer is specific: the abstract shows that the test changed detection and reports a stage IV secondary signal, while missing its primary endpoint. It does not establish that offering multi-cancer blood screening helps people live longer—and it does not settle whether future follow-up will show such a benefit.