Colonoscopy Reduces Risk of Death from Colorectal Cancer in High-Risk Patients
Long-term results from a study of colonoscopy for patients at higher-than-average risk of colorectal cancer confirm that removing precancerous adenomas can not only reduce the risk of colorectal cancer but also reduce the number of deaths from the disease by more than half, according to this study. The findings appeared February 23 in the New England Journal of Medicine.
These results, from the National Polyp Study, may not apply to the general, average-risk population. However, the findings provide reassurance that removing precancerous adenomas decreases the risk of death from colorectal cancer in people at higher-than average risk, noted lead author Dr. Ann Zauber of Memorial Sloan-Kettering Cancer Center.
“In screening, what you’re going to pick up for the most part is adenomas, not cancers, so I think it’s really important to know the impact [on mortality] of taking out those adenomas when you do a colonoscopy,” said Dr. Zauber. (The benefit of colorectal cancer screening for average-risk individuals is being tested in several randomized trials. See the sidebar.)
Earlier results from the National Polyp Study showed that colonoscopy and removal of adenomas found during the procedure was associated with a reduced incidence of colorectal cancer. However, whether the reduction in incidence found in the study would translate into fewer deaths could be determined only with longer follow-up.
If the study hadn’t shown a reduction in mortality, explained Dr. Zauber, that would have indicated that colonoscopy mostly picks up adenomas that would not progress to the aggressive cancers that lead to death from the disease. Instead, the finding that colonoscopy and adenoma removal reduced deaths from colorectal cancer by more than half over two decades of follow-up indicates that at least some of the adenomas that were detected would have progressed to cancer if they had not been removed.
Substantial Reduction in Deaths
In 1980, Principal Investigator Dr. Sidney Winawer of Memorial Sloan-Kettering and his co-investigators began the National Polyp Study at seven clinical centers. The study was designed to determine the appropriate interval for follow-up colonoscopy after adenoma removal.
Of the more than 9,000 original participants, 3,778 had at least one polyp removed; of those, 2,632 had adenomas. The other 776 patients, with benign polyps, served as an internal control group for the follow-up portion of the study, to track survival in people who did not have any adenomas at the time of initial colonoscopy.
The median follow-up for all patients was almost 16 years. The researchers compared the number of deaths from colorectal cancer in the group that had adenomas removed with the number of deaths in the internal control group without adenomas. They also compared the number of deaths in the adenoma group with the number of deaths expected to have occurred in the general population, calculated from NCI’s Surveillance, Epidemiology, and End Results database, the National Center for Health Statistics database, and the National Death Index.
Over the follow-up period, 12 patients in the adenoma group died of colorectal cancer, compared with an expected 25.4 deaths from the disease in the general population. This translates to an estimated 53 percent reduction in the risk of death from colorectal cancer following the removal of precancerous adenomas during colonoscopy.
Only one patient in the nonadenoma group died of colorectal cancer during the follow-up period. In the first 10 years after initial colonoscopy, the risk of death from colorectal cancer was approximately the same between the patients who had precancerous adenomas removed and those without adenomas. “We felt that was a powerful result,” said Dr. Zauber.
Modeling the Natural History of Adenoma Removal
To determine the actual mortality benefit of adenoma removal, the ideal randomized trial would compare colorectal cancer mortality in a group of patients who have had adenomas removed with that in a group of patients who have had adenomas detected but not removed, explained Dr. Asad Umar, chief of the Gastrointestinal and Other Cancers Research Group in NCI’s Division of Cancer Prevention (DCP). But, he added, such a study would be unethical because the evidence is clear that between 10 and 24 percent of adenomas will progress to cancer.
Because this type of comparison could never be performed, Dr. Zauber and her colleagues also used a computer model in their study called MISCAN-Colon to estimate the number of deaths that would have occurred in a population of the same age and with the same number of detected adenomas as the National Polyp Study group, but who did not have those adenomas removed.
The model, created by the NCI-sponsored Cancer Intervention and Surveillance Modeling Network (CISNET), uses validated data on the natural progression of adenomas to cancer in order to calculate the risk of death from colorectal cancer in a hypothetical population.
Based on MISCAN-Colon, the researchers estimated that 145 patients whose adenomas had not been removed would have died, whereas only 12 patients whose adenomas were removed died of colorectal cancer in the study. This model suggests a 92 percent reduction in mortality from colonoscopy and adenoma removal in this group of high-risk patients.
To establish optimal surveillance strategies for the general population, other studies will be needed to determine the best ways to identify patients with adenomas that are at low risk versus high risk of progression to colorectal cancer, said Dr. Zauber. Recent studies have shown that many high-risk patients do not receive adequate follow-up, whereas many low-risk patients receive too many surveillance colonoscopies. “We need to use surveillance judiciously because colonoscopy does have risks,” she added.
She also believes that colonoscopy quality and standards need to be improved. “We really think that our high-quality baseline colonoscopy [in the National Polyp Study] contributed to the strong reduction in both incidence and mortality that we observed,” she added.
For now, this study “adds evidence to support the widespread assumption—that I think is correct—that colonoscopy decreases the risk of colorectal cancer and decreases colorectal cancer mortality,” said Dr. Barry Kramer, director of DCP.
The 50 percent reduction in risk observed in this study may not apply outside this particular study population. "The participants were healthier overall than the average population, which could have lowered their risk of death, but they harbored adenomatous polyps, which could have increased their risk of death," explained Dr. Kramer.
“While this study was not randomized, and it is very difficult to match a study cohort using population data, this study confirms qualitatively that colonoscopy decreases mortality from colorectal cancer,” said Dr. Kramer. “However, we still can’t answer the question with precision: In the [general] public, what would be the absolute decrease in colorectal cancer mortality?”
Mature results of several large population-based colorectal cancer screening trials will not be available until after 2020, he added, although some preliminary results indicate that screening colonoscopy may reduce colorectal cancer mortality in the general public.