As many as 6% of colorectal cancers are discovered between 6 and 60 months of colonoscopy – usually in the proximal colon, and usually in patients whose index colonoscopy revealed adenomas, according to Dr. N. Jewel Samadder and colleagues.
While the investigators speculated that "differences in the biology of interval versus detected cancers" might be to blame, "suboptimal management of precancerous lesions (such as incomplete polypectomy) likely also plays a role," they wrote. The report is in the April issue of Gastroenterology (doi:10.1053/j.gastro.2014.01.013).
Dr. Samadder of the University of Utah, Salt Lake City, and colleagues looked at data from 126,851 patients who underwent colonoscopies between 1995 and 2009. The two centers in this analysis serve more than 85% of the state’s population. Patients with a history of colorectal cancer before 1995 were excluded, the researchers wrote.
Overall, 2,659 patients were diagnosed with colorectal cancer during the study period, mostly at initial, index colonoscopy – the investigators referred to these as "detected" cancers.
Dr. N. Jewel Samadder
However, 159 of these cases (6%) were diagnosed between 6 and 60 months of the index colonoscopy, and were therefore classified as "interval" cancers.
Interval cancer patients had a mean age of 67 years at index colonoscopy (range, 34-92 years), identical to that of detected cancer patients; both cohorts were split evenly between men and women.
The authors found that compared with detected cancers, the interval cancers were much more commonly discovered in the proximal colon – indeed, proximal cancers made up 55% of the cohort, with the remainder split evenly between the distal colon and the recto-rectosigmoid junction.
In comparison, only 39% of detected cancers were found in the proximal colon (P less than .001).
Patients with an adenoma or villous adenoma discovered at index colonoscopy were significantly more likely to develop interval cancer, compared with patients whose cancers were detected earlier (for adenoma, the odds ratio was 2.96, 95% confidence interval = 2.0-4.28; for villous adenoma, OR = 2.04, 95% CI = 1.34-3.11).
A family history also conferred an odds ratio of 2.27 for interval cancer, compared with detected cancers (95% CI, 1.24-4.16).
Finally, interval cancers were more likely to be associated with polypectomy or a biopsy at index colonoscopy (84.3%), compared with patients who did not undergo these procedures (51.8%).
According to the authors, "the association of proximal tumor location, earlier stage at diagnosis, and survival advantage compared with detected colorectal cancers suggests that tumor biology may play an important role in the pathogenesis of these lesions."
On the other hand, a 2010 study showing a much higher interval cancer rate suggests that inexpert polypectomy may be at least partially to blame, since most colonoscopies in that study were performed by nongastroenterologists (Am. J. Gastroenterol. 2010;105:2588-96).
The National Cancer Institute, the American Society for Gastrointestinal Endoscopy, the American College of Gastroenterology, the Huntsman Cancer Foundation, the Utah Department of Health, and the University of Utah funded the study. Dr Samadder and a coauthor disclosed consulting for Cook Medical, Covidien, and Myriad Genetics.