The NHS-Galleri trial randomised 142,250 people aged 50 to 77 to a blood-based multi-cancer early detection test added to usual care, or to usual care alone, and this paper reports prespecified secondary test-performance endpoints across three annual rounds in the intervention arm. Positive results were returned for 0.80% to 1.03% of participants per round, with 937 cancers detected by the test in total, episode sensitivity of 26.7% to 37.2% for all cancers and 47.6% to 63.4% for twelve prespecified types, and specificity of 99.50% to 99.60%. Positive predictive value fell across rounds from 58.0% to 45.8%.
Why it is interesting: It measures what a multi-cancer blood test picks up in a randomised population programme rather than in case-control cohorts, and how that changes after the first round.