GRAIL's Galleri multi-cancer blood test missed the primary endpoint of its NHS trial, undermining the FDA approval case and dropping the company's shares more than 20% in 2026.1
The three-year NHS-Galleri trial reported top-line results on February 1, 2026.1 The primary endpoint — a statistically significant reduction in late-stage cancer detection — was not met. That result is the crux of the problem: proving earlier detection translates into population-level clinical benefit is harder than showing the test can find cancer signals in blood.
Insurance reimbursement now looks unlikely in the near term without FDA clearance. Galleri sales continue growing, but primarily through out-of-pocket purchases — a commercially narrower path that limits market scale and restricts access to patients who can afford to pay directly.1
GRAIL management will present detailed NHS-Galleri trial data at ASCO 2026.1 That presentation is critical. Secondary endpoints — such as stage shift in specific cancer types — could support a revised regulatory argument. If the data doesn't deliver, the commercial strategy shifts further toward a premium, cash-pay niche.
Galleri uses AI and machine learning to analyze cell-free DNA fragments in the bloodstream, screening for signals from dozens of cancer types simultaneously. The clinical promise is real: detecting pancreatic, ovarian, and other hard-to-catch cancers before symptoms appear could shift survival outcomes. The NHS trial did not disprove that promise, but it failed to confirm it at the population level required for regulatory and reimbursement approval.
The trial outcome exposes a persistent challenge in multi-cancer early detection (MCED): lead-time bias, follow-up protocol variation, and patient compliance all complicate mortality benefit data. Detecting cancer earlier does not automatically translate into catching fewer late-stage cases in a randomized trial.
Peer liquid biopsy companies pursuing FDA approval are watching closely. GRAIL's regulatory timeline now sets a precedent — and a warning — for the entire MCED sector.
The validation cycle for AI-driven MCED platforms is running longer than anticipated.1 The financial pressure of that delay is real, as is the structural risk of building a business on out-of-pocket sales rather than the scale that insurer coverage would deliver. ASCO 2026 is not just a scientific presentation — it is a commercial inflection point.