It is widely accepted that the best way to determine whether screening can reduce deaths from cancer is with a randomized controlled trial (RCT), where half of a large group of people are invited to get a screening test, and scientists measure whether their risk of death is decreased compared to the half who did not receive an invitation to get screened. When it comes to colorectal cancer, we have that kind of evidence for fecal occult blood testing (FOBT, or stool testing) and for flexible sigmoidoscopy. But there has never been an RCT for colonoscopy, although several are now underway. The evidence supporting the value of colonoscopy has largely been based on inference.
A pair of new studies in this week’s New England Journal of Medicine provide new evidence supporting the value of colonoscopy in reducing colorectal cancer deaths, principally as a result of the detection and removal of precancerous polyps. We asked Robert A. Smith, PhD, senior director of cancer screening to comment on the significance of the new data.
“A persistent question related to colorectal cancer screening has been, ‘which test is best?’ The American Cancer Society colorectal cancer screening guidelines currently endorse seven different screening tests, including three stool tests, two endoscopic tests, and two radiographic tests. Each has been shown to be effective, but it is difficult to say one test is better than the other.
“The most recent ACS guidelines prioritize colorectal cancer screening tests that are more likely to identify adenomatous polyps, which are known precursors of colorectal cancer. Tests that visualize the colon, i.e., endoscopic tests and radiographic tests are more likely to detect polyps compared with stool tests.
“Two studies in this week’s New England Journal of Medicine help fill in some blanks. The first, by Zauber and colleagues at Memorial Sloan-Kettering Cancer Center, finds that removing pre-cancerous polyps during colonoscopy ultimately resulted in reduced colorectal cancer incidence, and eventually 53% fewer colorectal cancer deaths compared with the expected death rate in the general population. Yet, modeling data show that what colonoscopy achieves in one visit, high sensitivity stool tests, such as the fecal immunochemical test (FIT), have a good chance of achieving with annual testing over a ten year period.
“Actual empirical data is preferable to modeling data, and the initial results from a prospective randomized trial in Spain in the same issue of the journal holds the promise of eventually providing a prospective comparison of colonoscopy and high sensitivity stool testing. In this study, one-time colonoscopy is being compared to receiving FIT every two years in a population of 53,000 adults ages 50-69. In the first round of screening, the rate of colorectal cancer detection was higher in the colonoscopy group compared with the FIT group, and the rate of advanced adenoma detection was twice as high. This is expected since test-sensitivity is higher in colonoscopy compared with stool testing. As the study proceeds, we would expect the FIT group to ‘catch up,’ and if our models have been correct, the performance of the two tests over a 10 year period should become more similar. The success of this head-to-head study depends on making sure the study’s participants continue getting biennial FIT testing, something rarely achieved outside of a study. Longer follow-up likely will be needed in the first study, as well, to measure whether the lifesaving benefits of removing precancerous polyps will persist.
“In an editorial accompanying the studies, independent scientists suggest that colonoscopy could be used to help determine an individual person’s risk, based on the presence or absence of polyps at an initial screening at age 60. The idea of hybrid screening, where adults of different genders and different risks begin and end screening at different ages and are screened with different technologies based on risk is not a new idea, but this idea is still in its infancy, so we’re a long ways from proposing a practical alternative to the current guidelines which recommend that average risk adults begin screening at age 50 with either colonoscopy every 10 years, or annual high-sensitivity stool tests.”