Only modest benefit of colon cancer screening with colonoscopy

A one-time colonoscopy for screening entailed an 18% risk reduction of incidence in colorectal cancer (CRC) compared with no screening.

The trial included 95,000 patients, with 42% taking a colonoscopy

The NordICC trial (NCT00883792) is the first, randomised, population-based trial to investigate the effect size that CRC screening by colonoscopy has on incidence and mortality of CRC1,2. This magnitude of effect is important not only on an individual base for shared decision making with a patient but also for planning official screening programmes.

95,000 healthy people between 55 and 64 years of age were randomised directly from population registries in the partaking countries of Norway, Poland, Sweden, and the Netherlands to either receive an invitation for 1 colonoscopy screening or usual care that according to Prof. Michael Bretthauer (University of Oslo, Norway) meant no screening in the trial countries. Of the invitees, 42% participated in a colonoscopy. The presented analysis of the 10-year results could not include the 9,780 patients from the Netherlands due to Dutch General Data Protection Regulations. 

“In terms of screening performance, it went very well: there was good bowel preparation for the vast majority of the patients, cecal intubation rate exceeded 95%,” Prof. Bretthauer commented on the participant’s characteristics. The screening colonoscopies did not lead to any deaths or perforations and all 15 cases of major bleeding could be managed endoscopically.

Results informative, but not as trustworthy as those from intention-to-treat population

In the intention-to-treat population, the result for the primary endpoint of CRC incidence after 10 years revealed a cumulative CRC risk of 1.20% in those without screening compared with 0.98% in those in the invited group. This resulted in a risk ratio (RR) of 0.82 (95% CI 0.70–0.93) or a risk reduction of 18% with screening. Prof. Bretthauer also presented the results of an adjusted per-protocol analysis, not without underlining that these results are not as trustworthy as those from the intention-to-treat population. “But they are still informative, because what you do, or try to do is, to estimate the effect if everybody had accepted the screening,” he explained.

The per-protocol assessment showed an incidence of 1.22% on usual care and 0.84% with colonoscopy screening, leading to a risk reduction of 31% (RR 0.69; 95% CI 0.55–0.83). As for the secondary endpoint of CRC mortality, the difference between groups was low: 0.31% (usual care) versus 0.28% (invited group) with a non-significant risk reduction of 10% (RR 0.90; 95% CI 0.64–1.16). In his summary, Prof. Bretthauer further added that with regard to all-cause mortality there was no difference between the 2 groups. 

  1. Bretthauer M, et al. Long-term colorectal cancer incidence and mortality after screening colonoscopy: the NordICC randomised trial. LB08, UEG Week 2022, Vienna, Austria, 8–11 October.
  2. Bretthauer M, et al. N Engl J Med. 2022 Oct 9. Doi: 10.1056/NEJMoa2208375.