A 22% reduction in Stage IV diagnoses sounds like a win — until you read that the trial’s own primary endpoint, combined Stage III/IV cancers across all screened cancers, showed no statistically significant reduction. That gap between a missed primary and a compelling secondary readout is the central tension in GRAIL‘s NHS-Galleri data, and it reshapes how regulators, payers, and clinicians will receive what is otherwise genuinely striking signal.
The secondary findings are not noise. By the third annual screening round, Stage IV diagnoses across 12 prespecified cancers dropped 26% — suggesting the test’s value compounds with repeated use rather than manifesting in a single prevalent screen. That pattern is mechanistically coherent: a multi-cancer early detection blood test scanning for ctDNA signal will systematically miss cancers that haven’t shed enough DNA yet, and the first round captures a disproportionate share of slow-growing, low-shedding tumors that inflate apparent early-stage detection without moving the mortality needle. The 16% increase in Stage I–II diagnoses and the 25% reduction in emergency-presentation diagnoses are the numbers that tell a clinical story about real-world impact — particularly that emergency figure, which directly implicates downstream cost and quality of life in a way that staging shifts alone do not.
The trial design problem is now GRAIL’s regulatory problem. A randomized controlled trial that misses its primary endpoint does not become a success because secondary endpoints perform. The FDA will want to see overall survival data before granting a label indication for population-level screening, and that data is years away. In the interim, GRAIL is arguing implicitly that stage shift — specifically Stage IV reduction — is a valid surrogate for survival benefit. That argument has precedent in oncology, but it requires the agency to accept that preventing a Stage IV diagnosis translates into prevented death, a link the trial was not powered to establish directly.
The one number worth tracking in the months after ASCO is whether CMS engages seriously with coverage analysis. Medicare’s National Coverage Determination process for multi-cancer early detection has been stalled precisely because no trial has delivered on a primary endpoint. The 26% Stage IV reduction in Round 3 gives GRAIL its strongest argument yet — but coverage without a primary endpoint hit will set a precedent the agency will not make lightly.
Jon Napitupulu is Director of Media Relations at The Clinical Trial Vanguard. Jon, a computer data scientist, focuses on the latest clinical trial industry news and trends.
