Polyp Resection - Controversial Practices and Unanswered Questions.

Detection and complete removal of precancerous neoplastic polyps are central to effective colorectal cancer screening. The prevalence of neoplastic polyps in the screening population in the United States is likely >50%. However, most persons with neoplastic polyps are never destined to develop cancer, and do not benefit for finding and removing polyps, and may only be harmed by the procedure. Further 70-80% of polyps are diminutive (≤5 mm) and such polyps almost never contain cancer. Given the questionable benefit, the high-cost and the potential risk changing our approach to the management of diminutive polyps is currently debated. Deemphasizing diminutive polyps and shifting our efforts to detection and complete removal of larger and higher-risk polyps deserves discussion and study. This article explores three controversies, and emerging concepts related to endoscopic polyp resection. First, we discuss challenges of optical resect-and-discard strategy and possible alternatives. Second, we review recent studies that support the use of cold snare resection for ≥5 mm polyps. Thirdly, we examine current evidence for prophylactic clipping after resection of large polyps.

INTRODUCTION

Endoscopic colorectal cancer (CRC) prevention reduces CRC mortality by ~50%.[1,2,3] The benefit of colonoscopy on cancer prevention is dependent on effective polyp detection and removal.[1] Three important aspects of our approach to managing colon polyps are in evolution:

The first aspect is related to the management of diminutive polyps (polyps up to 5 mm in size). Because diminutive polyps are very common and almost never contain cancer, new management strategies to improve cost-effectiveness have been proposed.[4, 5] The “resect-and-discard” strategy that uses real-time polyp diagnosis of diminutive polyps has been recently endorsed by endoscopy societies.[5] However, adoption of this strategy into clinical practice faces several challenges.

The second topic is related to resection of mid-size polyps.[6] While “hot” electrocautery snare resection has been the standard-of-care for several decades, recent studies suggest that “cold” snare resection without electrocautery may be as safe and effective as hot snare resection.[6, 7]

Finally, several studies have focused on the management of large ≥20 mm polyps, particularly with respect to lowering the risk of bleeding complications.[8, 9] Prophylactic clip closure of the mucosal defect has become a common practice; however, whether clipping truly decreases bleeding risk remains unclear.

In this article, we discuss recent research developments and controversies with regards to the management of diminutive polyps, cold snare resection of mid-size polyps, and bleeding prophylaxis after resection of large polyps. The discussed observations call for refocusing our cancer prevention efforts in practice and research from removing diminutive polyps to the detection and safe resection of higher-risk polyps.

Should we adopt a resect-and-discard strategy for diminutive polyps?

Current colonoscopy practice guidelines recommend to remove, whenever possible, all polypoid lesions for histopathology assessment irrespective of the size or appearance. Of all detected polyps, 70–80% are diminutive, and ~50% of diminutive polyps are non-neoplastic. Cancer is exceedingly rare, and previous studies have described cancer prevalence between 0 and 0.08% for diminutive polyps and cancer prevalence between 0 and 1.5% in polyps up to 10 mm.[10, 11, 12, 13, 14, 15, 16, 17] A recent and largest cross-sectional study to date included >42,000 polypectomies of up to 9 mm polyps did not find any cancer in any of these diminutive or small polyps.[18] Resection of diminutive polyps increases patient risk and cost, yet the benefit on cancer prevention by removing diminutive polyps is questionable.[15, 19, 20] However, histopathology evaluation of diminutive polyps remains important because presence of adenoma may determine low- or high-risk status of the patient and affect the colonoscopy surveillance interval.[21]

One avenue to reduce colonoscopy related cost would be to replace histopathology assessment by using endoscopic image, enhancing modalities to distinguish neoplastic from non-neoplastic polyps (Figure 1).[5, 22] Novel image modalities have shown to predict neoplastic polyps with high accuracy and thus allow to determine the interval for the subsequent surveillance colonoscopy.[23][24] This new ability has inspired the concept of the “resect-and-discard” strategy. According to this strategy, diminutive polyps are diagnosed real time as adenomas or non-adenomatous polyps by using digital chromoendoscopy, like narrow band imaging (NBI). Polyps that are diagnosed with high confidence are resected and discarded, while others are sent for pathology evaluation. Calculated cost savings of this approach have been estimated to be 33,000,000$ per year in the United States.[19] A >90% agreement between the optical and the pathology-based surveillance recommendations has been set as the required quality benchmark in order to adopt the resect-and-discard strategy.[5] In recent years, multiple studies have shown that this benchmark can be accomplished, but only if optical diagnosis is done with high confidence by experienced endoscopists in an academic setting.[25, 26] On the basis of these results, the resect-and-discard strategy has been endorsed by the European and American Societies for Gastrointestinal Endoscopy (ESGE and ASGE).[27, 28, 29, 30] Both societies are well aware of challenges with successful implementation of resect-and-discard. It requires training, credentialing and monitoring of quality. Legal aspects of discarding tissue may also be a concern. Further, the resect-and-discard approach requires additional efforts by the endoscopist during the examination, added photo and text documentation, and the need to combine optical with pathology based diagnoses in a large proportion of patients.[31] This added complexity to everyday practice might further hinder widespread adoption. Alternative concepts to simplify the resect-and-discard strategy and minimize or eliminate the need for optical and histopathology assessment have recently been proposed.[32] In a post hoc analysis, a non-optical resect-and-discard strategy was examined, in which all rectosigmoid diminutive polyps were considered as hyperplastic and all polyps proximal to the rectosigmoid as neoplastic.[33] The non-optical strategy agreed with the pathology-based surveillance recommendations in 89% of patients, just shy of the 90% benchmark, but not significantly different from the optical strategy. The non-optical strategy also reduced the number of required pathology examinations and provided more patients with surveillance recommendations immediately following the colonoscopy compared to the optical resect-and-discard strategy. While resect-and-discard is a promising idea to reduce colonoscopy associated cost, further research on how to make the concept feasible for community practice is warranted.

Figure 1

Diminutive polyps examined with white light and optical chromoendoscopy (a) adenomatous polyp in white light imaging (b) adenomatous polyp in optical chromoendoscopy (c) hyperplastic polyp in white light imaging (d) hyperplastic polyp in in optical chromoendoscopy.

The low risk of diminutive polyps to develop cancer might support an even more radical approach to polyp management, namely to defer resection of diminutive polyp, and to only remove those that have grown to higher-risk polyps during the surveillance interval. Ignoring diminutive polyps is already an accepted CRC screening practice with CT colonography.[34, 35, 36, 37] According to CT colonography guidelines, polyps ≤5 mm are not reported and exams repeated every 5 years.[35, 37] Available studies on natural history of polyps, albeit few support this approach. The risk of transition to cancer increases with size, and cost-effectiveness studies have only considered that polyps larger than 5 mm would transition to cancer.[38] In two follow-up CT colonography studies, the majority of 6-9 mm polyps (65–78%) did not grow within 2–3 years.[39, 40] Interestingly, approximately one quarter decreased in size. Only a small proportion grew and none progressed to cancer among a total of 401 polyps. When considering even smaller ≤5 mm polyps, it is plausible that their resection does not sufficiently contribute to CRC prevention to justify their removal, and its associated risk and cost.[41, 42] Instead, overdiagnosis and overtreatment may be a concern.

Although leaving diminutive polyps in place would constitute a paradigm shift, in a recent survey 72% of gastroenterologists would be agreeable to leave diminutive polyps in place if such an approach was endorsed by governing societies.[43, 44] Also, in daily practice, gastroenterologists may not resect diminutive polyps when their appearance suggests non-adenomatous tissue.[43] It has been estimated that deferring resection of diminutive polyps would result in a 64% reduction of therapeutic interventions during colonoscopies.[44] Prospective studies will have to show the safety and efficacy of this approach and whether it truly does not affect overall effectiveness of screening. In addition, patients' expectations and fears with regard to perceived cancer risk of deferring polyp resection would have to be evaluated.

Should we use cold snare resection for all polyps?

Polyps are removed either by forceps or by snare. While both are comparable for ≤3 mm polyps, larger polyps are insufficiently removed with a forceps and should be resected with a snare.[45, 46] In current practice, the most common approach to removing medium and large sized polyps ≥5 mm is to use electrocautery or “hot” snare resection.[43, 47, 48] Added cautery ablates marginal tissue and may therefore improve completeness of resection (). Further, it might lower the risk of immediate bleeding. However, there is little proof that these assertions are true.

Hot snare resection is often incomplete. In the complete adenoma resection (CARE), study 10% of 5–20 mm neoplastic polyps were not completely removed. Incomplete resection increased with size and varied broadly across endoscopists.[6] A recent smaller study from Australia presented in abstract form reported an only 5% incomplete resection of up to 20 mm polyps when using a cold snare suggesting that resection may be at least be similarly complete when using a cold snare than a hot snare.[49]

With respect to bleeding risk a randomized trial among patients on anticoagulation with up to 10 mm polyps suggests that cold snare resection may actually lower the risk of bleeding.[50] Cold snare resection resulted in a lower-immediate (6 vs. 23%) and delayed bleeding risk (0 vs. 14%) when compared to hot snare resection. Two other randomized trials compared cold to hot snare resection for up to 8 mm polyps. In one study, no immediate or delayed bleeding occured.[46] The other study reported immediate bleeding in 9% with cold snare resection; however, all resolved spontaneously, and none required an intervention.[51] These studies are small and bleeding is not well defined, however, the results question the assumed benefit of cautery on bleeding risk. Aside from randomized trials, an increasing number of uncontrolled cohort studies suggest that cold snare polypectomy is safe and effective for up to 10 mm polyps.[46, 50, 52]

Current commonly used snares have been designed to be used with electrocautery and may not easily cut through the polyp base without cautery. Specialized fine wire snares to facilitate cold resection have been introduced.[52, 53] A first randomized trial compared the use of a dedicated cold snare with a standard snare for cold resection of up to 10 mm polyps and found a lower incomplete resection rate with using a dedicated cold snare (9 vs. 21%).[52]

While there are increasing number of studies on the safety and efficacy of cold snare resection for polyps up to 10 mm, the data on cold snare resection for larger polyps are limited to feasibility. Case series have reported on piecemeal resection of up to >20 mm colorectal polyps.[7, 49, 54] Among those three studies, immediate bleeding requiring intervention only occurred in one patient who was on anticoagulation treatment.

Finally, it should be noted, that cold snare polypectomy has no risk of cautery damage to the colonic wall or the resected polyp. This absence of electrocautery may decrease complications (perforation, post-polypectomy syndrome) and allow for better histopathology evaluation of the polyp and examination of the resection margins after polyp removal. Remnant polyp tissue may be more readily visible, which may improve completeness of resection.

Although the data on cold snare resection of larger polyps is still emerging, cold snare resection appears to be at least as safe and effective as hot snare resection for polyps up to 10 mm in size. Future studies should systematically examine efficacy and safety of cold snare resection; particularly determine the upper limit for en-bloc and piecemeal resection, the need for submucosal injection, and associated bleeding risks.[15]

Should we use prophylactic clipping after large polyp resection?

The risk of advanced histology of transition to cancer increases with polyp size. The prevalence of cancer in polyps equal or larger than 10 mm has been reported to be between 2.4 and 10.2%.[9, 10, 11, 12, 14, 15, 16] Therefore, it is of vital importance to assure complete resection particularly of large lesions. The standard-of-care for large polyps used to be surgical resection. A growing number of studies have demonstrated that resection can effectively and safely remove 85–90% of these polyps and has therefore become the preferred treatment.[9, 55, 56, 57]

For non-pedunculated polyps, endoscopic mucosal resection (EMR) is the current standard-of-care in Western countries. En bloc resection is the goal because the risk of recurrence is lower when compared to piecemeal resection.[58] However, in the majority of these polyps en bloc resection cannot be achieved and are therefore removed piece-meal. EMR typically includes submucosal injection with a contrasting agent (methylene blue or indigo carmine) to provide a submucosal safety cushion and to better delineate the submucosal layer. With the application of electrocautery to a larger area of the colonic wall during resection and the resultant large mucosal defect, the risk for perforation, post-polypectomy syndrome, and bleeding increases. It does therefore not surprise that the risk for complications increases with lesion size. It ranges overall from 8 to 26% in prospective studies.[8, 9, 55, 56, 59, 60]

The most common complication, delayed bleeding, is observed in 3–10% of patients.[9, 55, 61, 62, 63, 64] Bleeding typically occurs within 7–10 days after EMR and may require hospital admission, endoscopic intervention, and blood transfusion. Proximal polyp location, size, and bleeding during the resection have been identified as risk factors for delayed bleeding.[9, 55, 62, 63, 64, 65] Age, comorbidities, use of anticoagulation, and electrocautery setting have been reported risks in some but not other studies.[63, 65, 66]

Recent publications suggest that closing the mucosal defect with clips may reduce the bleeding risk (Figure 3).[63, 65, 67] A large retrospective single endoscopist study found that complete clip closure after removal of ≥20 mm polyps among 225 patients was associated with a 2% bleeding risk, significantly lower than the 10% observed in 247 historical controls, who did not undergo prophylactic clipping.[63] Similarly, clip closure was associated with lower bleeding risk in a recent prospective multicenter cohort study from Spain.[65] Although both studies support the use of clips, a major concern is an uncontrolled study design with the possibility of patient selection bias and unmeasured factors that may affect bleeding risk. Although the study from Spain was large (1214 patients), the lack of a standardized resection protocol, a clear definition of outcomes and assurance of complete outcome assessment are some of the limitations. For instance, clip closure may have been more frequently applied to lower-risk lesions because they may have been easier to clip (preferential clipping), which may have confounded risk assessment. A randomized trial from China might have overcome these limitations. Prophylactic clip closure was compared to no clip closure after endoscopic resection of ≥10 mm sessile polyps among 348 patients.[67] Polyps were removed either by EMR, endoscopic submucosal dissection (ESD) or hybrid ESD (combination of EMR and ESD). Delayed bleeding occurred less frequently after clip closure compared to no clipping (1 vs. 7%). However, the study included smaller polyps, allowed different endoscopic resection techniques, observed a higher than expected rate of complications in the control group, and used an unclear definition of bleeding events.[67] Therefore, the results are not sufficient to inform current EMR practice.

Figure 3

EMR of a 40 mm lateral spreading granular type colon polyp (a–c). Treatment of the boarders with APC to prevent recurrence (d) and clip closure of the mucosal defect after resection (e,f).

Despite the lack of good evidence, several observational studies indicated that clip closure is increasingly applied. Additional studies are under way and will hopefully provide a more definitive answer in the near future. At present stage, however, there is no adequate evidence to support prophylactic use of clips after EMR.

SUMMARY

Adequate polyp management is key for effective endoscopic CRC screening. While we spent most of our time and resources on the detection and removal of diminutive polyps, it is unclear that this effort is worthwhile. The proposed resect-and-discard strategy is an approach to shift this balance. However, training, monitoring, auditing requirements, challenges in implementation, and the added complexity may further hinder adoption into clinical practice. Alternative strategies include a simplified resect-and-discard strategy or deferring removal of diminutive polyps until they grow to higher-risk polyps. New strategies need to be studied, particularly with a focus on trade-offs of safety compared to cost-savings.

As we may de-emphasize the importance of removing diminutive polyps, our effort should have a renewed focus on the detection and complete resection of higher-risk polyps. Cold snare resection may be at least as safe and effective as hot snare resection for polyps up to 10 mm in size. Future comparative effectiveness studies should be encouraged to define best practice. The larger the polyp the higher the risk for complications. Although clipping of the mucosal defect after resection seems to be increasingly performed, there is currently insufficient evidence to support this practice for all non-pedunculated ≥20 mm polyps. Results of ongoing studies are awaited to understand if this approach is justified.

Study Highlights

✓ Adequate polyp management is key for colonoscopy practitioners. New developments like the resect-anddiscard strategy, deferring removal of diminutive polyps and considerations how polypectomy techniques can be improved are discussed in this article.