SAN DIEGO – Screening colonoscopy showed its efficacy for preventing incident cases of colorectal cancer in prospectively collected data during follow-up of up to 24 years in about 170,000 average-risk Americans.
The finding adds prospectively collected data from a large database of average-risk Americans to the evidence supporting routine colonoscopy screening for colorectal cancer. In contrast, data from the influential National Polyp Study assessed screening colonoscopy in high-risk patients who first underwent polypectomy (N. Engl. J. Med. 2012;366:687-96), Dr. Paul Lochhead said at the annual Digestive Disease Week.
Dr. Paul Lochhead
In general, the new findings support current public health recommendations for screening colonoscopy every 10 years in adults aged 50-75 years, but with a few caveats.
The new results showed that screening colonoscopy can significantly cut the risk for new-onset colorectal cancer by 51%, that the benefit from a single colonoscopy screen extended beyond 7 years, and that colonoscopy worked better than screening sigmoidoscopy. The findings also highlighted that colonoscopy was much better at preventing new distal cancers compared with its efficacy for stopping incident proximal tumors, and that when people had two, three, or more screening colonoscopies over time their risk of incident colorectal cancer fell further than after a single screening, said Dr. Lochhead, a gastroenterologist at the University of Aberdeen (Scotland).
"Although a single negative colonoscopy is associated with risk reduction, continued screening may be associated with greater benefit," undercutting the notion that average-risk middle-aged adults should undergo just a single screening colonoscopy with no follow-up screening if the first proves negative, he said.
In addition, "efforts to improve prevention [resulting from] proximal screening are warranted," he added.
"For distal cancers, we saw benefit beyond 10 years, but for proximal cancers we’re less certain about the duration of benefit. It appears there was a pattern of additional risk reduction with multiple screens regardless of whether the cancers were proximal or distal. That is something to bear in mind before we say that once is enough," he said in an interview.
The study used data from two large, prospective, U.S. observational studies: the Nurses’ Health Study, which began in 1976 and initially included 121,700 U.S. women, and the Health Professionals Follow-Up Study, which began in 1986 and included 51,529 men.
In the Nurses’ Health Study, data became available on screening endoscopy starting in 1984, so the data that Dr. Lochhead and his associates used through 2008 included up to 24 years of follow-up. The Health Professionals Follow-Up Study started tracking screening endoscopy in 1988, which gave as many as 20 years of follow-up data through 2008.
For this analysis, the researchers excluded participants in the two studies who had a lower endoscopy prior to the index procedures included in the two databases, those who had prior polyps or any cancer other than nonmelanoma skin cancer before they had their index endoscopy, and participants with inflammatory bowel disease. They calculated multivariate Cox proportional hazard models on the follow-up data that controlled for age, body mass index, smoking, family history of colorectal cancer, regular aspirin use, physical activity, diet, and multivitamin use.
The database included more than 2 million person-years of follow-up from both studies, and during follow-up 2,198 participants developed new-onset colorectal cancer. The majority of participants, constituting nearly 1.4 million person-years of follow-up, underwent no screening endoscopy during the years studied. About 424,000 person years of follow-up came after screening sigmoidoscopy, nearly 300,000 person years of follow-up came following screening colonoscopy, and over 65,000 person-years of follow-up came after polypectomy.
In the multivariate model, compared with no endoscopy, a negative colonoscopy result cut the rate of colorectal cancer by a statistically significant 51%, while negative sigmoidoscopy and polypectomy each cut the subsequent cancer rates by a statistically significant 37%, Dr. Lochhead reported.
When broken down by cancer site – proximal or distal – a negative colonoscopy was the only procedure to cut the risk for incident proximal cancers significantly, reducing the rate by 26% compared with no endoscopy. For distal cancers, colonoscopy cut the rate by 71%, compared with no screening, while sigmoidoscopy and polypectomy each cut the rate by 53%; all these risk reductions for distal cancers were statistically significant.
In the analysis that assessed the durability of protection, screening colonoscopy cut the risk for new colorectal cancers by a statistically significant 34%, compared with no screening, even when incident cancers were tallied more than 7 years following the index colonoscopy procedure. When cancers were divided by location, however, colonoscopy only provided significant protection for the first 3 years, with a risk reduction of 41% compared with no screening. Beyond that, colonoscopy did not produce a statistically significant reduction in incident cancers compared with no screening.