SAN DIEGO – A negative screening colonoscopy dramatically reduces the subsequent risk of colorectal cancer, according to a systematic review and meta-analysis of 18 studies presented at the annual meeting of the American College of Gastroenterology.

In the analysis, average-risk individuals whose colonoscopy showed neither cancer nor polyps had an incidence rate of colorectal cancer (CRC) of 0.58 per 1,000 person-years, corresponding to an estimated 10-year risk of just 0.58%, reported first author Dr. Larissa L. Fujii, a physician with the Mayo Clinic in Scottsdale, Ariz.

Compared with the expected risk for the general population based on surveillance data, these individuals were 57% less likely to receive a CRC diagnosis.

"Our findings support the effectiveness of colonoscopy as a screening and risk-stratification tool, and can be used to help educate patients who come in asking about their risk of colorectal cancer after having a negative colonoscopy," she commented.

Session comoderator Dr. Michael B. Wallace of the Mayo Clinic, Jacksonville, Fla., noted that a recent study found a higher incidence of CRC after a colonoscopy with polypectomy, at about 1.5 per 1,000 person-years of follow-up (Gut 2013 June 21 [Epub ahead of print]). "Do you think that your study might underestimate [the rate] because of follow-up? Was that a limitation in these patients?"

Adequacy of follow-up was included when assessing the studies’ methodologic quality, Dr. Fujii replied. "The incidence was lower in the higher-methodological-quality studies, so I would say that we are probably not underestimating. In fact, we might still be overestimating the risk based off of the population and the quality subgroup analyses."

Dr. Wallace also noted that there was a much lower CRC rate after negative colonoscopy for studies conducted in hospital settings as compared with those conducted in population settings. "Which do you think might be more accurate in terms of the rate?" he asked.

"I do think, because the population probably captured more colorectal cancers, that might be more indicative of what the actual incidence rate is," Dr. Fujii said.

Session attendee Dr. Douglas Robertson of the Geisel School of Medicine at Dartmouth, Hanover, N.H., said, "We like to think we are doing better colonoscopy over time. So can you see any differences as you look at newer studies versus older studies?"

The investigators have not yet assessed temporal trends, but identifying any might be difficult because only a single study was conducted before 2000, according to Dr. Fujii.

"I guess the subtext here is, is 10 years the right interval" for repeating colonoscopy after a negative result? Dr. Robertson further asked. "So you have thought a lot about this. What do you think – is a 57% reduction enough to stick with the 10-year interval?"

"I do like to think that this does support the 10-year recommendation – repeat the colonoscopy after 10 years rather than doing what many people might do, which is 5 years for average-risk patients, because the incidence is lower than what we would expect for the general population," Dr. Fujii replied.

In the study, the investigators searched for cohort studies and randomized controlled trials of average-risk patients undergoing screening colonoscopy. Studies were included if they had a mean follow-up duration exceeding 1 year.

All of the 18 studies meeting inclusion criteria were cohort studies, according to Dr. Fujii.

Thirteen studies each were conducted in hospital-based settings and in Western countries. The patients included in the studies had a weighted mean age of 72.5 years, and the weighted mean percentage of males was 41%.

The mean duration of follow-up was 5 years (range, 2-12 years). During follow-up, nearly 7,000 colorectal cancers were diagnosed.

Main results showed that the pooled incidence of CRC after a negative colonoscopy was 0.058% per year.

In subgroup analyses, the rate was significantly lower in studies conducted in hospital- versus population-based settings (0.08 vs. 0.96 per 1,000 person-years) and in studies conducted in Eastern versus Western countries (0.05 vs. 0.66). The rate also differed significantly according to whether studies were of high, moderate, or low quality (0.01, 0.88, and 0.38, respectively).

However, there was no significant difference according to whether the physician performing the colonoscopy was a gastroenterologist or some other specialist.

The estimated 5- and 10-year risks of CRC after a negative colonoscopy were 0.29% and 0.58%, respectively. These values compared with the expected rates of 0.6% and 1.5% for the general population according to Surveillance, Epidemiology, and End Results (SEER) data.

The difference amounted to more than halving of the rate of CRC after a negative colonoscopy relative to the general population (rate ratio, 0.43).

Dr. Fujii disclosed no relevant conflicts of interest.

Risk of CRC sharply lower after negative colonoscopy