Results from the largest randomized trial of a multicancer blood test found that adding GRAIL's Galleri test to usual care did not significantly reduce the incidence of late-stage cancers — the study's primary endpoint — but several secondary findings appeared more encouraging. 

Primary results from the NHS-Galleri trial, which randomized more than 142,000 asymptomatic adults in England, showed no statistically significant decrease in combined stage III and IV cancers among people screened for cancer with Galleri's multicancer blood test plus usual care vs usual care alone. But secondary analyses suggested stage IV diagnoses declined in later screening rounds, early-stage detection increased, and more cancers were detected through screening across successive rounds.

While GRAIL investigators argued the findings suggest the test can shift cancers toward potentially more treatable stages, several independent experts urged caution in interpreting the results.

Eric Klein, MD, distinguished scientist at GRAIL, emphasized the secondary findings, explaining that the trial showed "a marked reduction in stage IV" and more cancers diagnosed at potentially curable stages.

But Richard Houlston, PhD, who provided outside expert commentary via the UK's Science Media Centre, argued that the researchers "presented their findings far more positively than the overall results justify." 

While Houlston acknowledged that some secondary findings are "encouraging," he called it "entirely speculative" to conclude that the increase in detected stage III cancers reflected earlier detection of cancers that may have otherwise presented at stage IV cancers.

Even if the trial's primary endpoint had been met, "there would still be valid criticism and uncertainty about the interpretation of the results, because this is not a mortality endpoint," said Nitzan Rosenfeld, PhD, director, Barts Cancer Institute, Queen Mary University of London, also via the Science Media Centre.

Findings were presented at the American Society of Clinical Oncology (ASCO) 2026, and published in the Journal of Clinical Oncology.

What the Trial Found, What It Means

The Galleri test, developed by GRAIL, analyzes patterns of DNA methylation in fragments of cell-free DNA circulating in the blood. Cancer cells shed DNA with distinctive methylation signatures, and the test uses machine learning to identify those patterns and predict a cancer signal of origin, narrowing the likely tumor site to guide diagnostic workup. 

The NHS-Galleri trial is the first randomized clinical trial to assess a multicancer early detection test. The trial randomized more than 142,000 adults aged 50 to 77 in eight cancer alliance regions in England to usual care plus annual Galleri testing or usual care alone. Participants with a positive test result entered standard National Health Service diagnostic pathways; those with a negative test returned for 1- and 2-year follow-up blood sample collections. 

The primary endpoint — a reduction in combined stage III and IV cancer incidence across 12 prespecified cancer types responsible for roughly two-thirds of cancer deaths — was not met (incidence rate ratio [IRR], 1.03; 95% CI, 0.92-1.14; P = .63). In the first screening round, stage III and IV cancers were 19% more common in the intervention than the control arm (IRR, 1.19), but this difference did not reach statistical significance (95% CI, 0.98-1.43). The authors hypothesized that an increase in stage III diagnoses may reflect earlier detection of cancers that would otherwise have been diagnosed at stage IV.

Secondary analyses showed stage IV diagnoses decreased by 22% in the second screening round (IRR, 0.78; 95% CI, 0.57-1.06) and 26% in the third (IRR, 0.74; 95% CI, 0.57-0.95), although only the third-round reduction reached statistical significance. Stage I and II diagnoses increased by 16% over three rounds of screening (relative risk 1.16; 95% CI, 1.03-1.30) — 647 in the Galleri arm and 559 in the usual care arm. 

Over three screening rounds, 1801 participants (0.91%) had positive signal for cancer and 937 of those (52%) were diagnosed with cancer. The test, however, missed roughly 70% of cancers diagnosed within 12 months of the blood draw.

Overall, Klein believes the findings indicate that the clinically meaningful divide is no longer between stage III-IV and stage I-II, but between stage IV and stage I-III — where curative intent therapy is possible. "If one is screened 3 consecutive years with Galleri, you're less likely to be diagnosed with metastatic cancer, and far more likely to be diagnosed with stage II or III cancer."

Rosenfeld cautioned that interpreting results after a primary endpoint miss carries "a risk that we might be 'moving the goal posts,'" calling the incident-round stage IV trend "the most exciting part of the results" but said it "remains an untested possibility" that would need longer follow-up to confirm.

A separate study published in JAMA suggested a modest, temporary increase in diagnostic delays for patients in the control group during the first year of the trial. This "spillover" effect could theoretically overestimate the value of Galleri testing within the trial, the authors and experts noted in an accompanying JAMA editorial. 

Experts also cautioned about the potential consequences of false-positive and false-negative test results. About half of positive tests did not lead to a cancer diagnosis, while the test missed roughly 70% of cancers diagnosed within a year of screening, which could mean "many people may go on to have additional scans or biopsies that ultimately do not find cancer," explained Houlston, head of the Division of Genetics and Epidemiology, Institute of Cancer Research, London.

If a patient is told their MCED result is negative, "are clinicians providing false reassurance, especially given the relatively low sensitivity for stage I-II cancers?" Richard Sullivan, MD, PhD, of King's College London, and Christopher Booth, MD, of Queen's University, Kingston, Canada, asked in the recent JAMA editorial.

Overall, Mark Robson, MD, of Memorial Sloan Kettering Cancer Center, New York City, emphasized that mortality remains the most important endpoint. "Increased diagnostic yield should not be the standard by which we judge screening tests," he explained during an ASCO discussion. Instead, the key question is whether the test saves lives — an endpoint the trial will assess over 6 years of follow-up.

GRAIL is pursuing FDA premarket approval based on test performance and safety data from the NHS-Galleri trial and 25,490 US participants in Pathfinder 2.

No major oncology society currently recommends MCED testing for routine population screening, and outside experts noted the latest findings are unlikely to change that. 

However, "the end of this story has not yet been written," according to Sullivan and Booth. Survival from the NHS-Galleri study is being evaluated, and multiple other MCED tests in development could prove useful for cancer screening as well as cost effective for health systems. "Finding the correct balance in this space will require scrupulous science, the right end points, and honest conversations that prioritize scientific humility and public good over hype," Sullivan and Booth wrote.

The NHS-Galleri trial was funded by GRAIL. Charles Swanton, MBPhD, who presented the study findings at ASCO, reported stock ownership in and a consulting/advisory role with GRAIL, among other disclosures. Klein is an employee of GRAIL. Robson reported no financial relationships with GRAIL. Sullivan, Booth, and Houlston reported no relevant financial relationships. Rosenfeld disclosed relationships with Inivata and NeoGenomics and inventor status on cancer-detection patents.