“The COLONPREV trial—published online in The Lancet—is, therefore, a landmark study. It is the first randomised controlled trial to compare colorectal cancer deaths in people screened with the two most commonly used methods: colonoscopy and faecal immunochemical test (FIT), an antibody-based test for haemoglobin, indicative of blood in the stool. The study finds that invitation to biennial FIT-based screening is non-inferior to invitation to one-time colonoscopy in terms of colorectal cancer mortality at 10 years. These results build on the NordICC trial, published in The New England Journal of Medicine, which showed that invitation to colonoscopy reduced colorectal cancer risk compared with no screening. Participation in screening was low in both trials but was higher for those assigned to FIT-based screening than colonoscopy in COLONPREV: 20·1% of those assigned to colonoscopy underwent colonoscopy, whereas 39·4% of those assigned to FIT took the stool-based test, although not all of these participants did so every 2 years. Participation in any form of screening was 31·8% in the colonoscopy arm and 39·9% in the FIT arm. The results support the wider adoption of FIT-based screening and suggest that such programmes improve participation. [..]
A major determinant of the decision to undergo screening is the invasiveness of the test on offer. People returning a positive FIT must still undergo more invasive colonoscopy. However, the negative predictive value of FIT is more than 99% and the COLONPREV study shows that about four in five people can safely avoid a colonoscopy with FIT as the primary screening test. It is possible that blood-based tests for colorectal cancer might be more acceptable than stool tests or colonoscopy, but current evidence suggests their effectiveness is insufficient for use in screening. To aid personal decision making regarding colorectal cancer screening—whichever test is offered—the public must be made aware of the benefits of screening, including the prevention of future cancers, as well as any potential harms. Clinicians must also be better equipped to have personalised discussions about cancer risk. Predictive models for colorectal cancer might help to guide such conversations and support risk-adapted screening, although clinical trials and evidence from real-world application of such tools are needed.”
Full editorial, The Lancet, 2025.3.29