Registration delays likely led to ‘major underestimation’ of CRC risk reduction in NordICC
Click Here to Manage Email Alerts
Key takeaways:
- NordICC had 221 post-randomization exclusions due to prior CRC diagnosis, suggesting delays in cancer registration.
- These delays likely underestimated 10-year CRC risk reduction and screening colonoscopy effects.
Delays in cancer registration may have underestimated 10-year colorectal cancer risk reductions by up to 75% in the NordICC trial, with researchers warning that study estimates on screening colonoscopy should be “interpreted with caution.”
“Epidemiological studies have long suggested strong effects of screening colonoscopy in preventing CRC,” Hermann Brenner, MD, MPH, of the division of clinical epidemiology and aging research at German Cancer Research Center, and colleagues wrote in JAMA Network Open. “However, much weaker than expected effects were reported in 2022 from the NordICC trial, the first and so far only randomized clinical trial reporting on long-term effects of screening colonoscopy, to our knowledge.”
The NordICC trial used national registries to identify 85,179 individuals aged 55 to 64 years without a CRC diagnosis in Norway, Sweden and Poland between 2009 and 2014, who were randomly assigned to receive or not receive an invitation for screening colonoscopy. Of those, 221 participants were later excluded due to a prior CRC diagnosis identified after randomization.
“In the NordICC trial, CRC was ascertained by record linkage with cancer registries,” Brenner and colleagues wrote. “Registration delays are a well-known common concern in population-based cancer registration.”
To estimate the impact of cancer registration delays, the researchers compared the number of post-randomization exclusions with expected CRC diagnoses and derived outcomes associated with these delays.
Brenner and colleagues calculated an expected 80 new diagnoses within 1 year in the NordICC trial, based on cancer registry data. However, the exclusion of 221 individuals who were not initially identified with CRC “suggests a 2- to 3-year delay in registration,” they wrote.
In addition, results from 10 years of follow-up show CRC risk differences of 0.22% and 0.38% in intention-to-screen and per-protocol analyses, respectively, which suggest that 455 people would need to be invited to screen and 263 people would need to complete screening colonoscopy to prevent one CRC event.
“The reported risk differences exclusively emerged in the 4-year period from 6 to 10 years of follow-up,” Brenner and colleagues wrote. “A mean registration delay by 1, 2 to 3 years would imply that only approximately 75%, 50% or 25% of individuals diagnosed in that 4-year period would have been registered.”
These delays suggest that the risk difference, as well as the true numbers needed to invite or undergo colonoscopy, were underestimated by approximately 25% with a 1-year delay, 50% with a 2-year delay and 75% with a 3-year delay.
“The findings of this study suggest that delay of cancer registration most likely led to major underestimation of 10-year CRC risk reduction in the analyses of the NordICC trial reported in 2022, in which ascertainment of incident cases had most likely remained incomplete for a large proportion of the participants recruited in 2009 to 2014,” Brenner and colleagues wrote. “Updated analyses ensuring complete 10- and 15-year follow-up will be crucial to derive the true reductions of CRC risk and mortality in the trial’s predefined interim and primary analysis.”
They continued: “In the meantime, available estimates are to be interpreted with caution, as they likely severely underestimate true screening colonoscopy effects.”
Collapse
Read more about
Click Here to Manage Email Alerts