Rishi Surana, MD, PhD
Kimmie Ng, MD, MPH
Colorectal cancer ranks third among cancer deaths in both men and women in the United States, with an estimated 150,000 new cases and 52,000 deaths anticipated in 2022.1 Colorectal cancer rates have declined by approximately 2% per year from 2014 to 2018 in people older than age 50, which is thought to be partly related to more widespread uptake of colorectal cancer screening. Because the best screening test is the one that gets done, the U.S. Preventive Services Task Force and other medical societies recommend several options for colorectal screening, including colonoscopy, flexible sigmoidoscopy, fecal occult blood testing, fecal immunochemical testing, and computed tomography colonography.
Of these modalities, only two have demonstrated a mortality benefit in randomized clinical trials, which are considered the gold standard method of determining the efficacy of a screening intervention: fecal occult blood testing and flexible sigmoidoscopy. A landmark randomized study conducted by the Minnesota Colon Cancer Control Study group compared colorectal cancer screening using fecal occult blood testing with a control group of subjects who did not undergo colorectal cancer screening; they found that annual fecal occult blood testing resulted in a 33% reduction in colorectal cancer–specific death compared with no screening.2 Similarly, a randomized controlled trial, conducted in Norway, of flexible sigmoidoscopy vs no screening demonstrated that flexible sigmoidoscopy reduced the rate of colorectal cancer–specific death by 27%.3
It is important to note that all patients with abnormal screening results in both studies were offered follow-up colonoscopy. Despite fecal occult blood testing and flexible sigmoidoscopy having the most rigorous supporting evidence, the primary modality of colorectal cancer screening offered to patients by primary care physicians and gastroenterologists in the United States is colonoscopy.
The NordICC Trial
Prior to the publication of the Nordic-European Initiative on Colorectal Cancer (NordICC) study by Bretthauer et al in TheNew England Journal of Medicine this year,4 there were no randomized clinical trial data of colonoscopy for colorectal cancer screening. The only previously available data supporting the use of colonoscopy coming from prospective cohort studies estimated an approximately 60% to 70% reduction in mortality associated with colonoscopy.5,6 However, cohort studies are notoriously rife with limitations such as the inability to address potential confounding variables. Thus, the NordICC trial, as the first published randomized clinical trial testing the efficacy of colonoscopy, fills an important knowledge gap and provides new insight into the real-world population-wide benefit of colonoscopy for colorectal cancer screening.
The NordICC trial was a randomized, pragmatic study that enrolled 84,585 adults (aged 55 to 64) from Poland, Norway, and Sweden.4 Participants were randomly assigned in a 1:2 ratio to the “intervention” group—a one-time invitation to undergo screening colonoscopy—or to the usual-care group who did not undergo any cancer screening. The trial was strategically performed in geographic areas where no organized colorectal cancer screening programs were available at the start of the trial, which ensured against contamination of the control arm of the study. The co-primary endpoints of the study were risks of colorectal cancer and death from the disease after a median follow-up of 10 and 15 years, with the 10-year endpoint data reported in the article.
“[T]he NordICC trial … fills an important knowledge gap and provides new insight into the real-world population-wide benefit of colonoscopy for colorectal cancer screening.”— RISHI SURANA, MD, PhD, AND KIMMIE NG, MD, MPH
Tweet this quote
The results showed that individuals in the invited group had a 0.98% risk of colorectal cancer at 10 years (95% confidence interval [CI] = 0.86–1.09) compared with 1.20% (95% CI = 1.10–1.29) in the usual-care group, with a risk ratio of 0.82 (95% CI = 0.70–0.93), meeting one of the two primary endpoints of the trial. Surprisingly, and perhaps soberingly, there was no statistically significant difference between colorectal cancer–specific death between the invited and usual-care groups (0.28% vs 0.31%), nor was there a difference in all-cause mortality between the groups (11.03% vs 11.04%).
Putting the Study Results Into Context
Given the lack of colorectal cancer–specific and all-cause mortality benefit, does this mean that colonoscopy is not a useful screening modality for colorectal cancer? The short answer is no. Although the investigators should be applauded for undertaking and implementing a thoughtful and well-designed study, the NordICC trial has important limitations that warrant a more nuanced discussion.
To begin, this was not a study to unequivocally determine the utility of colonoscopy for colorectal cancer screening—this was a study to evaluate the efficacy of implementing a population-based screening program. Among the individuals randomly assigned to the invited group, screening was performed in just 42%, which is well below the current colorectal cancer screening adherence rate of 67% among age-eligible adults in the United States, with 60.6% of those people undergoing colonoscopy as the screening method.7
Although it is fair to conclude from the NordICC trial that a population-based screening program utilizing a one-time invitation for colorectal cancer screening does not work to realize the full protective benefits of cancer screening, we cannot conclude from NordICC that colonoscopy is ineffective as a screening tool. Indeed, a prespecified, per-protocol analysis in the NordICC study suggested that among the 42% of individuals who actually underwent screening colonoscopy, the 10-year risk of colorectal cancer decreased from 1.22% to 0.84% (which equates to a 31% risk reduction), and colorectal cancer–specific death decreased from 0.30% to 0.15% (a 50% risk reduction). Although per-protocol analyses are understandably controversial, they at least provide proof of concept that the intention-to-screen analysis in the NordICC study may underestimate the true magnitude of benefit of screening colonoscopy.
Dissecting the Data
Although the per-protocol analysis suggests that colonoscopies do lead to meaningful decreases in colorectal cancer incidence and mortality, the risk reductions were admittedly lower than expected based on previous literature. One possible reason is that mortality benefits may take a much longer time to manifest due to the long natural history of colorectal cancer; thus, the 15-year follow-up data will be critically important.
Another potential explanation for the more modest results seen in NordICC may lie in the quality of the endoscopic exams performed, since colonoscopies are highly operator-dependent and vary with the skill level and procedure volume of the endoscopist. Although the cecal intubation rate was quite good (> 95%), the adenoma-detection rate—defined as the proportion of colonoscopies where an adenoma is detected—was suboptimal.
An adenoma-detection rate of 25% is a commonly accepted benchmark, but previously published data from this group showed that 28.6% of the endoscopists had an adenoma-detection rate below the 25% threshold.8 So, it is possible that operator-related differences were present and may have affected the optimal detection of premalignant and malignant lesions that could have impacted outcomes in the study.
Another critical question to consider is whether the two study arms in the trial were truly similar in terms of baseline participant characteristics. Although randomization usually mitigates imbalances between treatment arms, the data provided by the study are limited, so it is unclear whether the two arms were comparable in all relevant factors, particularly with respect to the underlying risk for developing colorectal cancer.
This is of particular concern when interpreting the per-protocol analysis, since it is unclear whether the participants in the invited arm who actually underwent colonoscopy may have been at higher risk for colorectal cancer and, thus, more willing to undergo screening compared with the people in the invited group who elected not to undergo screening. Any such difference in risk profile would not be reflected in the per-protocol analysis and could bias subsequent calculation of colorectal cancer incidence and mortality endpoints.
Preventing Colorectal Cancer
Finally, it is important to highlight an advantage of colorectal cancer screening that distinguishes it from all other cancer screening modalities. Although low-dose computed tomography for lung cancer screening, mammography for breast cancer screening, and prostate-specific antigen testing for prostate cancer screening aim to diagnose early-stage disease, colorectal cancer screening with endoscopy is unique in that it aims to, in part, prevent disease by identifying and removing premalignant lesions.
“[C]olorectal cancer can be a preventable disease, but only if people actually undergo screening.”— RISHI SURANA, MD, PhD, AND KIMMIE NG, MD, MPH
Tweet this quote
Although the ultimate endpoint for evaluating the utility of a cancer screening modality is usually considered to be disease-specific and all-cause mortality, these outcomes do not tell the whole story with colorectal cancer screening. A diagnostic colonoscopy can certainly identify early-stage disease, but its ability also to completely prevent the development of colorectal cancer leads to enormous personal, medical, and societal benefits by eliminating the morbidity associated with a colorectal cancer diagnosis and its subsequent treatment.
The 18% relative risk reduction in colorectal cancer in the invited group in the NordICC trial may seem modest at first glance. However, the downstream cumulative benefit of preventing colorectal cancer in this proportion of the population is substantial, including eliminating the need for surgery, chemotherapy, radiation, and frequent utilization of radiologic exams and health-care resources, as well as the anxiety associated with a cancer diagnosis. We eagerly await the release of additional data such as rates of hemicolectomy, chemotherapy administration, and quality-of-life assessments from the NordICC trial investigators.
Concluding Thoughts
The NordICC study has taught us that screening colonoscopies are safe and significantly reduce the incidence of colorectal cancer. Although the lack of impact on colorectal cancer–specific and all-cause mortality in the overall study population was disappointing, advances in the treatment of this cancer, particularly with respect to adjuvant therapies, could help explain this lack of mortality benefit, as well as insufficient follow-up time. Regardless, the decreased incidence of colorectal cancer with screening colonoscopies reported in the NordICC study is not trivial, as a diagnosis of the disease, even in early stages, can impose significant morbidity. In light of these trial results, should we still be recommending colonoscopy for colorectal cancer screening?
We believe the answer is yes. But, we should also conduct more nuanced discussions with our patients regarding the data supporting this screening and perhaps adopt a level of humility regarding the true benefit of screening colonoscopy. There are certainly important barriers to colonoscopy, including an inconvenient bowel prep, conscious sedation, need for transportation and time off work, and cost, and we now know from the NordICC trial that a simple invitation to undergo screening does not work.
How, then, can we overcome these challenges and increase the number of patients getting screened? For patients who are not able or willing to undergo screening colonoscopy upfront, we should encourage the use of other acceptable screening options, including flexible sigmoidoscopy or fecal occult blood testing, with the understanding that a colonoscopy is indicated for any positive test. Ongoing studies comparing screening colonoscopies with these other colorectal cancer screening modalities will help adjudicate the true value of screening colonoscopies in our armamentarium of tools.
And, finally, we need to rethink public health strategies and determine the changes that need to be made in our health-care systems to (1) improve the uptake of colorectal cancer screening, as the 42% in the NordICC trial and the 67% adherence rate to such screening in the United States are likely insufficient to realize its full benefits; and (2) achieve equitable access to high-quality screening across all segments of the population, particularly those who are most vulnerable and at highest risk for the disease.
After all, colorectal cancer can be a preventable disease, but only if people actually undergo screening.
DISCLOSURE: Dr. Surana and Dr. Ng reported no conflicts of interest.
REFERENCES
1. Siegel RL, Miller KD, Fuchs HE, et al: Cancer statistics 2022. CA Cancer J Clin 72:7-33, 2022.
2. Mandel JS, Bond JH, Church TR, et al: Reducing mortality from colorectal cancer by screening for fecal occult blood: Minnesota Colon Cancer Control Study. N Engl J Med 328:1365-1371, 1993.
3. Holme O, Loberg M, Kalager M, et al: Effect of flexible sigmoidoscopy screening on colorectal cancer incidence and mortality: A randomized clinical trial. JAMA 312:606-615, 2014.
4. Bretthauer M, Loberg M, Wieszczy P, et al: Effect of colonoscopy screening on risks of colorectal cancer and related death. N Engl J Med 387:1547-1556, 2022.
5. Nishihara R, Wu K, Lochhead P, et al:Long-term colorectal-cancer incidence and mortality after lower endoscopy. N Engl J Med 369:1095-1105, 2013.
6. Doubeni CA, Corley DA, Quinn VP, et al: Effectiveness of screening colonoscopy in reducing the risk of death from right and left colon cancer: A large community-based study. Gut 67:291-298, 2018.
7. QuickStats: Percentage of adults aged 50-75 years who met colorectal cancer screening recommendations: National Health Interview Survey, United States, 2018. MMWR Morb Mortal Wkly Rep 69:314, 2020.
8. Bretthauer M, Kaminski MF, Loberg M, et al: Population-based colonoscopy screening for colorectal cancer: A randomized clinical trial. JAMA Intern Med 176:894-902, 2016.
Dr. Surana is a physician in the Gastrointestinal Cancer Center at Dana-Farber Cancer Institute, and Instructor in Medicine at Harvard Medical School. Dr. Ng is Associate Chief, Division of Gastrointestinal Oncology; Director, Young-Onset Colorectal Cancer; and Co-Director, Colon and Rectal Cancer Center at Dana-Farber Cancer Institute and Associate Professor of Medicine at Harvard Medical School, Boston.
Disclaimer: This commentary represents the views of the author and may not necessarily reflect the views of ASCO or The ASCO Post.
