This is half of a two-part debate — read the opposing argument. The opinions expressed here are that of the author and do not represent the position of AGA.
The past decade has witnessed numerous technical and technological advances in the field of colonoscopy. One of the most important developments, however, has been the realization that colonoscopy is a powerful tool to prevent colorectal cancer (CRC) if it is done by good hands. This operator dependency is closely associated with the risk of post-colonoscopy CRC, as shown in the study by Corley et al.1 where each 1 percent increase in the adenoma detection rate (ADR) was associated with a 3 percent decrease in the risk of cancer.
There has been considerable interest in developing technologies which could help bridge the gap between high and low polyp detectors, and an overview of this ever expanding literature reveals some notable trends. New technologies can be expensive to acquire and maintain, and somewhat unwieldy to operate, hindering widespread adoption outside of specialized centers. An arguably more sobering observation is that while new technologies can work well and do improve colorectal neoplasia yield, the magnitude of the increment is variable and simply does not match the improvement observed when the colonoscope is taken from a low detector and given to a high detector.2 In other words, technology is helpful but cannot supersede technique fundamentals.
Enter distal colonoscope attachment devices, which I will argue are “good technique extenders,” at a fraction of the cost and possible inconvenience of more sophisticated devices and platforms. Several such accessories are currently available, including the transparent cap, Endocuff, and Endorings; all are based on the principle that maximizing colonic mucosa exposure requires effective and efficient flattening and separating of colonic folds. Endocuff accomplishes this with thin flexible projections, while Endorings utilizes sequentially spaced silicone disks; both also help prevent slippage during maneuvers such as loop reduction and interventions.
It is important to understand that distal colonoscope accessories can make good become better, but cannot make bad become good.
The obvious next questions are: do they work, and if so, which one is the best? The answer to the first question is a nuanced yes, and here is why. A review of the published literature comparing distal accessories to standard colonoscopy may frustrate seekers of unequivocal answers, as randomized trials and meta-analyses of these trials have shown inconsistent and sometimes contradictory results. A network meta-analysis3 (NMA) helps provide perspective. NMA is an analytical approach which can assess the comparative effectiveness of multiple interventions across a network of randomized controlled trials, and is well-suited to compare the relative performance of distal colonoscope accessories. This NMA included 25 randomized trials (more than 16,000 patients) which compared standard colonoscopy to the cap in 14, Endocuff in nine, Endorings in one and cap versus Endocuff in one. The primary outcome was ADR, and endoscopists’ baseline performance was defined as low (ADR 10 percent) or high (ADR 40 percent). The use of any device compared to none was associated with increased ADR (39.3 percent vs. 35.1 percent, relative risk (RR), 1.13; 95 percent CI, 1.03-1.23); however, the benefit was more marked for high performers whose ADR increased to 45.2 percent, whereas low performers’ ADR increased to 11.3 percent. Both Endocuff and Endorings were associated with increased ADR overall when each was compared to standard colonoscopy (Endocuff 40.4 percent vs. 34.6 percent, RR 1.21; 1.03-1.41, Endorings 49.1 percent vs. 28.8 percent, RR 1.70; 1.05-2.76), but not the cap (37.0 percent vs. 34.3 percent, RR 1.07; 0.96-1.19). Endocuff was associated with the most striking improvement among high detectors, whose ADR increased to 48 percent; however, improvements among low detectors were very modest. Secondary outcome analyses showed improved cecal intubation time (but not rate) for the cap and Endocuff, and a very favorable safety profile for all three devices. The NMA was inconclusive regarding the superiority of any one device due to the low quality of the evidence and paucity of head-tohead comparisons.
In my practice, I tend to use the cap and Endocuff the most, but for different reasons. I favor the cap for difficult polypectomies and endoscopic mucosal resections, especially in the case of flat polyps located on the proximal side of folds when stable positioning and effective mucosal exposure are critical to ensure complete resection. I also use the cap when performing a colonoscopy in a patient with suspected diverticular hemorrhage, because my ability to identify and treat a culprit diverticulum is anecdotally better. I use the Endocuff for more routine cases, not primarily to improve my ADR, but to make polyp detection more efficient; I find that I can flatten folds and expose mucosa more efficiently and with less effort and maneuvering than with standard colonoscopy. An ongoing randomized clinical trial (RCT) at our institution will help provide a more quantitative — and objective — assessment of these observations.
In the meantime, what are practical take-home messages? First, it is important to understand that distal colonoscope accessories can make good become better, but cannot make bad become good. This is illustrated by the NMA3 where devices helped high baseline detectors more than low detectors; in essence, supporting the principle that there is no substitute to rigorous fundamental technique and compulsive examination methods! Second, I would not scoff at the relatively modest improvements in ADR with distal accessories, notwithstanding that the increment is mostly due to small and diminutive lesions; suffice to remember that even numerically small gains in ADR are associated with decreased risk of post colonoscopy CRC. Third, distal attachment devices may help some endoscopists more than others, and this may be related to personal preferences and intangible biases as much as objective performance measurements. For example, in the RCT4 by Pohl et al. comparing cap-assisted to standard colonoscopy, there was no significant difference in adenoma detection, but endoscopists’ ADR varied from +20 percent to -15 percent with use of the cap, and this was associated with “individual preference,” but not low baseline ADR. Given that distal colonoscope devices do not come with a significant downside (i.e. easy to use, reasonable cost and excellent safety profile), I would encourage my colleagues and readers to familiarize themselves and try them in their practice, measure their effect on polyp detection… and, as I am, await more comparative data.
Dr. Kahi has no conflicts to disclose. Dr. Kahi is a Special Section Editor for Clinical Gastroenterology and Hepatology.
1. Corley, D.A., Jensen, C.D., Marks, A.R. et al, Adenoma detection rate and risk of colorectal cancer and death. N Engl J Med. 2014; 2014; 370: 2541.
2. Rex, D.K. Optimal withdrawal and examination in colonoscopy. Gastroenterol Clin North Am. 2013; 42: 429-42.
3. Facciorusso, A., Del Prete, V., Buccino, R.V. et al, Comparative Efficacy of Colonoscope Distal Attachment Devices in Increasing Rates of Adenoma Detection: A Network Meta-analysis. 2017; DOI: 10.1016/j.cgh.2017.11.007.
4. Pohl, H., Bensen, S.P., Toor, A. et al, Cap-assisted colonoscopy and detection of Adenomatous Polyps (CAP) study: a randomized trial. Endoscopy. 2015; 47: 891-7.