Galleri trial evidence supports earlier cancer detection despite unmet primary endpoint

A large clinical evidence program supports the Galleri multi-cancer early detection test’s ability to find clinically significant cancers at earlier stages, according to a correspondence arguing that earlier warnings about the test were overlooked. The authors, Eric A. Klein, Oliver Venn and Eric T. Fung, say studies involving nearly 400,000 participants indicate that Galleri can detect cancers before symptoms develop, often when curative treatment may remain possible. They also report a low false-positive rate and an acceptable safety profile in prospective studies. However, the correspondence acknowledges that the randomized NHS-Galleri trial did not meet its primary endpoint of reducing combined stage III and IV diagnoses for 12 specified cancers, and it states that a mortality benefit has not yet been established.

The central biological argument is that cancer detectability depends on more than tumor volume. The TRACERx study cited by the authors found that the extent to which lung tumors shed circulating tumor DNA is an intrinsic feature of tumor biology, with greater shedding associated with lymphovascular invasion and a high Ki67 proliferation index, both indicators of aggressive disease. Galleri does not search only for individual mutated DNA fragments. Instead, it analyzes genome-wide methylation patterns at thousands of informative CpG sites, which the authors say can produce a detectable signal even when mutant allele fractions are extremely low. They also cite previous work suggesting that circulating tumor allele fraction is a stronger predictor of classifier performance than absolute ctDNA concentration or clinical stage alone.

The authors further contend that Galleri preferentially identifies tumors with a likelihood of causing harm rather than indiscriminately detecting every small growth. In substudy 3 of the Circulating Cell-free Genome Atlas, five-year survival among participants whose tumors were detected by Galleri was comparable to survival among people whose cancers were found through conventional clinical pathways. Participants whose tumors were not detected by the test generally had significantly better five-year overall survival than those whose tumors were detected. These observations are consistent with a test that tends to find biologically consequential cancers, although the correspondence does not present them as proof that Galleri can predict whether every individual tumor will become dangerous.

Diagnostic sensitivity in the early CCGA substudy rose with stage: 16.8% for stage I cancer, 40.4% for stage II, 77.0% for stage III and 90.1% for stage IV, with the respective 95% confidence intervals reported in the article. More recent prospective results came from PATHFINDER 2, in which slightly more than half of newly diagnosed cancers were found at stage I or II and about 70% at stages I through III. Nearly three-quarters of cancers detected by Galleri in that study occurred in cancer types without an existing recommended screening program. For 12 aggressive cancers accounting for roughly two-thirds of US cancer deaths, episode sensitivity reached 70%. NHS-Galleri produced results the authors describe as similar, reinforcing the evidence that the test can supplement screening for cancers for which routine population screening is unavailable.

Those results must be considered alongside the unsuccessful primary endpoint in NHS-Galleri. The trial did not find a statistically significant reduction in combined stage III and IV diagnoses among 12 prespecified cancers. Nevertheless, the authors report several secondary findings after three screening rounds: a 16% increase in stage I and II diagnoses, a reduction of more than 20% in metastatic diagnoses after rounds two and three, a fourfold increase in screen-detected cancers and a 21% reduction in diagnoses made during emergency presentation. Galleri detected 366 stage I or II cancers, compared with 290 cancers of all stages found through standard-of-care screening alone. These are observed trial outcomes associated with adding Galleri to standard screening, but they do not establish that the test reduced cancer mortality.

The correspondence disputes the view that the absence of a demonstrated mortality benefit makes a screening test clinically ineffective. It argues that clinical utility can also include earlier diagnosis, a greater opportunity for curative treatment, fewer metastatic presentations, more efficient diagnostic evaluation and reduced reliance on emergency care. The authors note that screening technologies have often been implemented through gradual evidence generation before definitive mortality benefits were available. They cite the recently reported NordICC colonoscopy trial as an example of a screening method that had not demonstrated a mortality benefit after 13 years of follow-up, although colonoscopy remains an accepted standard in the United States. Even so, NHS-Galleri was neither designed nor statistically powered to establish mortality reduction, making a definitive answer dependent on larger samples and longer follow-up.

The authors also reject the characterization of reductions in stage IV disease as merely a migration from one advanced stage to another. They point to improved five-year survival across US cancers, now exceeding 70%, with particularly large gains among patients diagnosed at regional stage. From their perspective, the clinically important distinction is whether a cancer is identified before it becomes metastatic, rather than relying on a simple division between early and late disease. The NHS-Galleri findings are consistent with that interpretation because metastatic diagnoses fell while earlier-stage diagnoses increased. The trial results nevertheless need longer follow-up to determine whether those shifts translate into lower cancer-specific deaths.

Prospective evidence provides a separate answer to concerns about false positives and procedural harm. PATHFINDER 2 reported a false-positive rate of 0.36%, a positive predictive value of 60%, cancer signal-of-origin accuracy of 91.7% and a median time to diagnostic resolution of 47 days. Fewer than 1% of participants underwent an invasive diagnostic procedure, and serious study-related adverse events were few. NHS-Galleri reported a false-positive rate of 0.45%, a positive predictive value of 58% and signal-of-origin accuracy of 92.5%, again with few serious study-related adverse events. Anxiety measured with validated instruments rose only mildly and transiently, returning to baseline within 12 months. The authors also cite an FDA advisory committee vote of 10–0 in favor of the test’s safety profile, while the present correspondence remains an argument by the test’s developers and does not substitute for independent regulatory review.

False negatives are interpreted as a limitation of Galleri’s intended role as an addition to, rather than a replacement for, established screening. A negative “No Cancer Signal Detected” result does not rule out cancer or justify discontinaging guideline-recommended screening. The authors say available observations do not support claims that a negative result reduces adherence: PATHFINDER 2 found no reduction at 12 months in colorectal, cervical, breast or lung cancer screening adherence, while real-world evidence found no reduction in mammography use during the 24 months after a negative result. PATHFINDER 2 also detected 91 cancers through standard screening and another 85 through routine care prompted by symptoms or incidental findings. NHS-Galleri separately detected 236 cancers through standard screening within its intervention arm, illustrating how the tests are intended to operate as complementary pathways.

Health-economic findings remain model-based rather than observed savings from the trial. Analyses cited in the correspondence estimate that Galleri used alongside standard care may be cost-effective over a lifetime horizon when changes in cancer stage, treatment intensity and diagnostic pathways are included, with the most favorable results when the high costs of advanced cancer care are incorporated. A separate model estimated that annual screening of approximately 13 million eligible people in England would increase colonoscopy demand by 2.09% and CT use by 0.76%, equivalent to about 1,040 additional colonoscopies and 4,720 CT examinations per million people screened each year. The correspondence also corrects the characterization of the Nancy Gardner Sewell Medicare MCED Screening Coverage Act: rather than mandating coverage, it creates a Medicare benefit category and gives the Centers for Medicare & Medicaid Services authority to apply its standard evidence-based review process. Taken together, the authors argue that continued evaluation is more justified than categorical dismissal, while mortality outcomes remain the unresolved question.

Subject of Research: Clinical and economic evidence for the Galleri multi-cancer early detection blood test

Article Title: Rebuttal to “Early Warnings About the Galleri Test Were Disregarded”

Article References: Klein, E. A., Venn, O., & Fung, E. T. (2026). Rebuttal to “Early Warnings About the Galleri Test Were Disregarded”. Clinical Proteomics, 23(1), Article 51. https://doi.org/10.1186/s12014-026-09644-x

Image Credits: AI Generated

DOI: 10.1186/s12014-026-09644-x

Keywords: Galleri, multi-cancer early detection, cancer screening, liquid biopsy, ctDNA, NHS-Galleri, PATHFINDER 2, false-positive rate, stage shift, cancer mortality, Rebuttal, Early