Real-world, population-based study
NordICC is an ongoing, pragmatic study and is the first randomized trial to quantify the possible benefit of colonoscopy screening on risk of colorectal cancer and related death.
Researchers recruited healthy men and women from registries in Poland, Norway, Sweden, and the Netherlands between 2009 and 2014. Most participants came from Poland (54,258), followed by Norway (26,411) and Sweden (3,646). Data from the Netherlands could not be included owing to data protection law.
At baseline, 84,585 participants aged 55-64 years were randomly assigned in a 1:2 ratio either to receive an invitation to undergo a single screening colonoscopy (28,220; invited) or to undergo usual care in each participant country (56,365; no invitation or screening).
Any colorectal cancer lesions detected were removed, whenever possible. The primary endpoints were the risks of colorectal cancer and colorectal cancer–related death. The secondary endpoint was death from any cause.
‘Modest effectiveness,’ but longer follow-up to give fuller picture
In an editorial that accompanied publication of the study, Jason A. Dominitz, MD, from the division of gastroenterology, University of Washington, Seattle, and Douglas J. Robertson, MD, from White River Junction (Vt.) Veterans Affairs Medical Center, commented on the possible reasons for the low reduction in incident cancer and deaths seen in NordICC.
They pointed out that cohort studies suggest a 40%-69% decrease in the incidence of colorectal cancer and a 29%-88% decrease in the risk of death with colonoscopy. However, they noted that “cohort studies probably overestimate the real-world effectiveness of colonoscopy because of the inability to adjust for important factors such as incomplete adherence to testing and the tendency of healthier persons to seek preventive care.”
Referring to Dr. Bretthauer’s point about attendance to screening, Dr. Dominitz and Dr. Robertson added that, in the United States, colonoscopy is the predominant form of screening for colorectal cancer and that in countries where colonoscopy is less established, participation may be very different.
“The actual effectiveness of colonoscopy in populations that are more accepting of colonoscopy could more closely resemble the effectiveness shown in the per-protocol analysis in this trial,” they wrote.
The editorialists also pointed out that the benefits of screening colonoscopy take time to be realized “because the incidence of colorectal cancer is initially increased when presymptomatic cancers are identified.” A repeat and final analysis of the NordICC data is due at 15 years’ follow-up.
In addition, they noted that “colonoscopy is highly operator dependent” and that the adenoma detection rate is variable and affects cancer risk and related mortality.
Given the “modest effectiveness” of screening colonoscopy in the trial, they asserted that, “if the trial truly represents the real-world performance of population-based screening colonoscopy, it might be hard to justify the risk and expense of this form of screening when simpler, less-invasive strategies (e.g., sigmoidoscopy and FIT [fecal immunochemical test]) are available.”
However, they also noted that “additional analyses, including longer follow up and results from other ongoing comparative effectiveness trials, will help us to fully understand the benefits of this test.”
Also commenting on the study was Michiel Maas, MD, from the department of gastroenterology and hepatology, Radboud UMC, Nijmegen, the Netherlands, told this news organization that he agreed that the absolute effect on colorectal cancer risk or colorectal cancer–related death was not as high as expected and may be disappointing.
But Dr. Maas said that “around half of the patients in the study did not undergo colonoscopy, which may have negatively impacted the results.
“An additional factor, which can be influential in colonoscopy studies, is the potential variability in detection rates between operators/endoscopists,” he said.
Looking to the future, Dr. Maas noted that “AI [artificial intelligence] or computer-aided detection can level this playing field in detection rates.
“Nevertheless, this is a very interesting study, which sheds a new light on the efficacy on screening colonoscopies,” he said.
Dr. Bretthauer has relationships with Paion, Cybernet, and the Norwegian Council of Research. Dr. Dominitz is cochair of VA Cooperative Studies Program #577: “Colonoscopy vs. Fecal Immunochemical Test (FIT) in Reducing Mortality from Colorectal Cancer” (the CONFIRM Study), which is funded by the Department of Veterans Affairs. Dr. Robertson is national cochair (with Dr. Dominitz) of the CONFIRM trial and has received personal fees from Freenome outside of the submitted work. Dr. Maas reported no relevant financial relationships.
A version of this article first appeared on Medscape.com.