FDA advisers vote on Galleri, the blood test that looks for many cancers at once

On 23 September 2026 FDA advisers voted on GRAIL's Galleri blood test for adults 50 and over. They voted 10 to 0 that it's safe, 6 to 4 that it works, and 7 to 2 (1 abstaining) that the benefits outweigh the risks. In the 2 big studies it caught about a third of cancers within a year (35.0% and 31.6%). The FDA hasn't decided yet. Rating: In clinical trials.

What is it?

Galleri is a blood test that looks for DNA shed by tumours. It counts as AI only because of its machine-learning classifiers: 3 locked computer models read the DNA patterns, decide whether there's a cancer signal, and guess where in the body it came from. A positive result needs follow-up tests; a negative one doesn't rule cancer out.

What's the evidence?

1. FDA's Molecular and Clinical Genetics Panel met on 23 September 2026 on GRAIL's premarket approval application. Votes: safety 10 yes, 0 no; effectiveness 6 yes, 4 no; benefits outweigh risks 7 yes, 2 no, 1 abstain. The panel's advice isn't binding on the FDA.^[2]

   Evidence stage: In clinical trials · #c1

2. The FDA's briefing says the test uses 3 locked, machine learning-based classifiers: one for whether a cancer signal is present, one predicting its origin (1 of 18 sites), and a supplemental one for some results.^[3]

   Evidence stage: In clinical trials · #c2

3. In PATHFINDER 2, the test picked up 35.0% of cancers diagnosed within 12 months; 77.0% of positive results turned out to be cancer and 99.85% of people without cancer correctly tested negative. In the NHS-Galleri trial the figures were 31.6%, 66.2% and 99.74%. In PATHFINDER 2 it found 21.4% of stage I cancers.^[3]

   Evidence stage: In clinical trials · #c3

4. The panel doubted the 'early detection' claim, said Galleri shouldn't replace usual screening, and wanted more studies after any approval, including on whether it improves patients' long-term outcomes.^[2]

   Evidence stage: In clinical trials · #c4

What does it mean for me?

Galleri is already sold in the US as a lab test (list price about $949), but it isn't FDA-approved, and the advisers' vote isn't a decision. If you're 50 or over and someone offers it, know that it missed about 2 in 3 cancers within a year, so it can't replace your usual bowel, breast, cervical or lung checks, or any prostate checks you've agreed with your doctor. About 1 in 4 positive results in the US study weren't cancer, which can mean worry and more scans or biopsies. No study has yet shown it helps people live longer.

What could I ask my doctor?

• If this test is negative, which of my usual screening tests do I still need?
• What happens next if it finds a cancer signal?

These are general starting points, not a recommendation to choose any particular test or treatment.

If you're worried about your health right now, contact your doctor or local health advice line. In an emergency, contact your local emergency number.

Change history

• 24 October 2026: First published.

General information only, not medical advice. Talk to your doctor about your situation. AI-assisted source check. Not clinically reviewed.