Endoscopic mucosal resection (EMR) is a continuously evolving technique that has revolutionized the diagnosis, staging, and treatment of superficial neoplasms throughout the digestive tract. As a minimally invasive technique, it offers an alternative to operative resection and mucosal ablation. For neoplasms limited to the mucosa, EMR compares equivalently with operative resection for R0 status (no residual neoplasia) and favorably for morbidity and mortality. In contrast with mucosal ablation, EMR delivers a more precise and predictable depth of tissue effect and, most important, provides the resected specimen for histopathologic analysis. EMR was pioneered in Eastern Asia as a technique for resection of early gastric cancers. Incrementally, the tools and techniques for EMR have evolved to be employed around the globe for eradication of superficial lesions throughout the esophagus, stomach, small bowel, and colon. Owing to improvements in endoscopic detection and recognition, more sessile, flat, depressed, and even subepithelial lesions are being identified for which endoscopic resection is being adopted. Moreover, the implementation of surveillance strategies for various gastrointestinal conditions, such as Barrett's esophagus, has fostered image guided detection and eradication of premalignant and early malignant lesions, incorporating EMR for diagnosis, staging, and therapy. This review focuses on the techniques, applications, and limitations of EMR for lesions throughout the digestive tract. EMR is an extension of standard snare polypectomy techniques for the eradication of epithelial-based lesions that are otherwise not amenable to standard resection. The intent is most often curative, but may also be used in selected instances for enhanced diagnostic and staging purposes. Other names for the EMR concept include mucosectomy and endoluminal resection. Considered broadly, EMR employs a variety of adjunctive tools and techniques to achieve en bloc or piecemeal resection, with the resected specimens made available for histopathologic assessment. Owing to the increased depth and volume compared with forceps biopsy tissue sampling, EMR specimens yield a more accurate diagnosis. When carcinoma is present, the EMR specimen provides important histopathologic prognosticators of tumor grade, depth of invasion, presence or absence of lymphovascular invasion, and tumor proximity to the deep and lateral resection margins. These features permit a determination as to the likelihood of R0 status versus the risk of residual mural tumor or concurrent lymph node metastases and affirm definitive endoscopic curative resection or a recommendation for operative or chemoradiotherapy. EMR is intended for curative resection of mucosal based lesions. Metaplasia (Barrett's epithelium) and dysplasia (adenomatous, squamous, low- and high-grade dysplasia), by definition, are mucosal based and suitable for curative EMR. Early mucosal cancers (T1a) have not penetrated beyond the lamina propria, have a negligible risk for concurrent lymph node metastasis, and are curable with complete EMR. Once carcinoma has penetrated into the submucosa (T1b), the risk of concurrent lymph node metastasis increases from up to 3% to 30%, depending on depth of invasion and tumor grade. As such, reliance on EMR for T1b lesions is highly individualized and dependent on the lesion and patient characteristics. Inasmuch it is undesirable to expose the patient to the risks of EMR without the benefit of curative resection, EMR should not be attempted in cases when there is a high likelihood of deep submucosal invasion unless EMR is being employed for diagnostic or staging purposes. Efforts at pre-resection discrimination between T1a versus T1b lesions is then of considerable clinical importance. Endoscopic ultrasonography has been used for locoregional staging to assess depth of lesion invasion and regional lymphadenopathy; however, the accuracy of endoscopic ultrasonography for submucosal invasion has more recently been questioned. Although indicated in selected instances, the routine use of endoscopic ultrasonography is being largely abandoned. Rather, an assessment of morphologic features, considerably improved with the adoption of high-definition and electronically enhanced imaging, is used to assess suitability for EMR. Frank ulceration and central depression predict invasive carcinoma. So, too, do immobility or fixation and failure to lift with attempt at submucosal injection or cap aspiration (Figure 1). The latter “non-lifting” sign may be falsely encountered when a prior incomplete attempt at resection has been performed with resultant fibroinflammatory scaring. Routine use of validated morphologic assessment schemes to predict likelihood of submucosal invasion would enhance selectivity (Figure 2). Submucosal invasion is more likely in Paris classification 0–IIa+c morphology, corresponding to a flat lesion with a depressed component, a nongranular surface, and a Kudo pit pattern type V, corresponding with amorphous, irregular pits. However, in the United States, gastrointestinal endoscopists have been slow to adopt such schema. Although endoscopists are notoriously inaccurate when it comes to lesion size estimates, size does matter. En bloc resection is suitable for most lesions ≤2 cm in diameter. Larger lesions typically require piecemeal resection. Single-session or incremental circumferential EMR can be successfully performed in the esophagus, stomach, and rectum; however, curative EMR is less likely in colonic and duodenal lesions that extend beyond 75% of the luminal circumference. Given the increased procedure-related time, risk, and skill set required, EMR is not advisable when initially discovered during a routine endoscopy. Rather, at the time of initial endoscopy, the lesion should be photographed and its location documented in reference to readily distinguishable landmarks. A limited number of cold forceps biopsies should be obtained without the use of electrosurgical energy to document the histopathology. Lesions deemed difficult to rediscover should be tattooed to facilitate identification in the future. Endoscopic tattooing should never be performed within or underneath the lesion of interest because the dye agent may result in a fibrotic response, making subsequent resection attempts more challenging. Once the lesion has been identified and the histology confirmed, the patient can be engaged in an informed discussion of the options for management including operative resection and dedicated endoscopic resection. The most commonly used EMR techniques are categorized as injection-, cap- and ligation-assisted EMR.1Kantsevoy S.V. Adler D.G. et al.ASGE Technology CommitteeEndoscopic mucosal resection and endoscopic submucosal dissection.Gastrointest Endosc. 2008; 68: 11-18Abstract Full Text Full Text PDF PubMed Scopus (266) Google Scholar Injection-assisted EMR (Figure 3), incorporates the injection of a solution into the submucosal layer beneath the lesion. This process lifts the mucosal-based lesion away from the muscularis propria, provides an elevated purchase about which a snare can be engaged, and reduces the electrosurgical resistance encountered during electrocautery snare resection. This constellation permits resection of broad areas of tissue while minimizing the risk of perforation and transmural burn syndrome. This is the most commonly used EMR technique for sessile lesions in the stomach, duodenum, and colon, where it is also referred to as “saline-assisted polypectomy.” Cap-assisted EMR employs a transparent cap that is fitted over the tip of the endoscope in combination with a crescent-shaped snare that is seated in the open position around the internal distal rim of the cap (Figure 4). After submucosal injection beneath the target lesion, the cap centered over the lesion and suction is used to retract the lesion up and into the cap. The snare is then closed to capture the neopolypoid tissue and suction is released. The ensnared tissue is then resected with application of electrocautery. Enthusiasts employ this technique for flat and nodular lesions in the esophagus and stomach. Concern over risk of full-thickness resection has limited its application in the colon and duodenum. Last, ligation-assisted EMR employs the use of a transparent cap affixed to the tip of the endoscope, modified with a band ligation system (Figure 5). Similar to the above, the tip of the endoscope is positioned over the target tissue, suction is applied and the lesion drawn up into the banding chamber. A band is deployed, entrapping the tissue and creating a neo-polyp. The neo-polyp is then resected using electrocautery snare technique applied below, above, or through the band. We and many others do not use pre-resection submucosal injection; however, some enthusiasts do so. There are several attractive aspects to this technique. First, the band routinely extrudes the muscularis propria layer entrapping only the mucosa and portions of the submucosa. Second, the system permits rapid and repeated applications to enable wide-area, confluent resections during a single session. This technique has become widely popular for the resection of flat and nodular dysplasia in the esophagus. The volume of fluid injected into the submucosal space varies based on size and location of the lesion and the fluid injected. Many solutions have been tested as submucosal injectates, and the ideal agent may yet to be determined. We routinely use methylene blue-tinted normal saline solution as our submucosal injectate. The addition of a chromic agent to the solution provides a contrast that better enables discrimination of the lesion's lateral margins and endoscopic assessment of completeness of resection. We are apt to add epinephrine to create a 1:10,000 solution for esophageal, gastric, and duodenal lesions to reduce “back-bleeding” from the needle insertion and acute bleeding post-resection. Normal saline solution is inexpensive and readily available; however, it quickly dissipates in some instances, making it necessary to incrementally repeat submucosal injections during piecemeal resection of larger lesions. Other agents intended to extend the duration of the submucosal injectate have included hyaluronic acid, hydroxypropyl methylcellulose, fibrinogen, glycerol solutions, and autologous blood. Overall, the performance characteristics of these agents have not proved to be sufficiently greater than normal saline-based solutions to promote widespread adoption. Although EMR is intended to achieve complete eradication of the entire target lesion, focal residual neoplasia may persist particularly after piecemeal resection of larger lesions and those that are difficult to access endoscopically. Adjunctive ablative therapies may enhance completion eradication of focal residua and decrease the risk of local recurrence after EMR. Argon plasma coagulation is a widely used ablative technique for completion eradication. Thermal ablation with the tip of the snare and bipolar or monopolar forceps have also been described. Esophageal EMR (Supplementary Video 1) was initially developed for curative resection of early moderately and well-differentiated squamous cell esophageal cancer and was found to have low complication rates with excellent 5-year survival outcomes, comparing favorably with esophagectomy. The Japanese Society for Gastrointestinal Endoscopy has suggested that EMR should be attempted for lesions with a diameter ≤2 cm and those involving less than one third of the circumference of the esophageal wall, although recent reports have demonstrated successful treatment of larger lesions, particularly in patients with high-grade dysplasia who are not considered operative candidates.2Ell C. May A. Gossner L. et al.Endoscopic mucosal resection of early cancer and high-grade dysplasia in Barrett's esophagus.Gastroenterology. 2000; 118: 670-677Abstract Full Text Full Text PDF PubMed Scopus (690) Google Scholar, 3Takeshita K. Tani M. Inoue H. et al.Endoscopic treatment of early oesophageal or gastric cancer.Gut. 1997; 40: 123-127PubMed Google Scholar With the increased prevalence and detection of Barrett's esophagus with high-grade dysplasia and/or intramucosal carcinoma, EMR is now widely used for patients with discrete nodules in an area of Barrett's mucosa for improved histologic diagnosis and staging, and is part of the armamentarium for Barrett's eradication therapy along with radiofrequency ablation and cryotherapy. The ability to safely and effectively curatively resect premalignant lesions and early mucosal cancer represents a dramatic shift from the prior management strategy of esophagectomy for those patients. Cap or ligation-assisted EMR is typically performed in the esophagus because the target lesions are typically flat or nodular. In selected patients, wide area, piecemeal ligation-assisted EMR as single-mode therapy successfully eradicated all high-grade dysplasia/intramucosal carcinoma and intestinal metaplasia in approximately 60% of patients with highly dysplastic Barrett's.4Bhat Y.M. Furth E.E. Brensinger C.M. et al.Endoscopic resection with ligation using a multi-band mucosectomy system in Barrett's esophagus with high-grade dysplasia and intramucosal carcinoma.Ther Adv Gastroenterol. 2009; 2: 323-330Crossref PubMed Scopus (15) Google Scholar EMR as single-mode therapy is best applied when there is short-segment and non-circumferential involvement. When combined with ablation therapy, complete eradication of all dysplasia and intestinal metaplasia can be expected in 90% of patients.5Ginsberg G.G. Endoscopic approaches to Barrett's oesophagus with high-grade dysplasia/early mucosal cancer.Best Pract Res Clin Gastroenterol. 2008; 22: 751-772Abstract Full Text Full Text PDF PubMed Scopus (7) Google Scholar For circumferential and long-segment Barrett's, the combined technique of EMR plus radiofrequency ablation is associated with a far lower complication rate than for that of radical, wide-area EMR as single-mode therapy.6Pouw R.E. Seewald S. Gondrie J.J. et al.Stepwise radical endoscopic resection for eradication of Barrett's oesophagus with early neoplasia in a cohort of 169 patients.Gut. 2010; 59: 1169-1177Crossref PubMed Scopus (188) Google Scholar The largest experience of gastric EMR has been described in Japan for resection of superficial early gastric cancers. Kojima et al7Kojima T. Parra-Blanco A. Takahashi H. et al.Outcome of endoscopic mucosal resection for early gastric cancer: review of the Japanese literature.Gastrointest Endosc. 1998; 48: 550-554Abstract Full Text Full Text PDF PubMed Google Scholar summarized the outcomes of 1832 patients from 12 major institutions who underwent EMR for early gastric cancers and found that complete resection was achieved in 73.9% of cases. Not surprisingly, lesions that were removed en bloc had a higher rate of complete resection compared with those resected in a piecemeal fashion. The overall combined complication rate was 1.9%. Subsequent studies have demonstrated that the local recurrence rate for early gastric cancers after EMR may be as high as 10%–20%. Sporadic, nonampullary duodenal adenomas can be successfully eradicated in 70% of cases with en bloc or piecemeal injection-assisted EMR.8Kedia P. Brensinger C. Ginsberg G. Endoscopic predictors of successful endoluminal eradication in sporadic duodenal adenomas and its acute complications.Gastrointest Endosc. 2010; 72: 1297-1301Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar Success rates are highest for lesions covering ≤50% of the luminal circumference. Inspection with a duodenoscope is important to ensure that the lesion does not arise for the ampulla of Vater (endoscopic papillectomy for ampullary adenomas is outside the scope of this review). We have observed higher rates of acute and delayed bleeding after EMR of laterally spreading duodenal adenomas such that we routinely add epinephrine (1:10,000) to our standard injectate. Unlike other EMR instances, we commonly admit patients for overnight observation after resection of duodenal lesions >2 cm or >25% of the luminal circumference because of the concern for post-procedure bleeding. With increased awareness, more flat and sessile colorectal neoplasms are being detected during screening and surveillance colonoscopy. In particular, there has been an increase in the detection and recognition of right-sided sessile serrated adenomas. Large, sessile, and flat adenomas are removed with en bloc or piecemeal injection-assisted EMR, including laterally spreading “carpet-like” lesions throughout the colon (Supplementary Video 2). Moss et al9Moss A. Bourke M.J. Williams S.J. et al.Endoscopic mucosal resection outcomes and prediction of submucosal cancer from advanced colonic mucosal neoplasia.Gastroenterology. 2011; 140: 1909-1918Abstract Full Text Full Text PDF PubMed Scopus (443) Google Scholar demonstrated that at tertiary referral centers in “expert hands” EMR with curative eradication was achieved in 89% of patients with sessile colonic polyps ≥20 mm in size. Independent predictors for failed EMR included previous intervention with electrosurgical energy, involvement of the ileocecal valve, and difficult position of the lesion. Lesion size ≥40 mm and the need for argon plasma coagulation for adjunctive ablation were considered independent predictors for adenoma recurrence. Bleeding is the most common complication observed with EMR. Although bleeding rates have varied in the literature, many larger series have demonstrated an average bleeding rate of ≤10%. A lack of standardization in reporting has contributed to our poor understanding of predictors and thereby a means to reduce acute and delayed bleeding rates. Acute bleeding from the resection site is managed with injection of dilute epinephrine solution, bipolar forceps, and/or hemostatic clips. Control of acute bleeding should permit completion of EMR. For practical purposes, we consider delayed bleeding as that which becomes clinically apparent after the patient has left the endoscopy facility. Delayed bleeding occurs anywhere from 12 hours to 12 days post EMR. Electrosurgical current selection for EMR has not been standardized and the optimal settings to reduce the risk of delayed bleeding have not been determined. For patients who are considered at high risk for post-EMR bleeding, including those who require anticoagulation after the procedure, we attempt prophylactic clip closure of the resection site. Warfarin and antiplatelet agents should be managed as per the ASGE guidelines.10Zuckerman M.J. Hirota W.K. Adler D.G. et al.ASGE guideline: the management of low-molecular-weight heparin and nonaspirin antiplatelet agents for endoscopic procedures.Gastrointest Endosc. 2005; 61: 189-194Abstract Full Text Full Text PDF PubMed Scopus (139) Google Scholar The rates of perforation have been reported to be 0.3%–0.5% with EMR, although the risk of perforation depends on the type of lesion, location, and display of non-lifting sign. Careful inspection of the EMR site and the resected specimen may allow the endoscopist to recognize acute perforation and allow for immediate closure with endoscopic clipping, thereby obviating the need for operative repair. Stricture formation is location dependent and is most apt to occur in the esophagus and duodenum. Strictures are more likely to develop with increased size of the target lesion. Esophageal and duodenal lesions that occupy >60%–75% of the luminal circumference are more likely to form strictures after EMR, whereas colonic lesions are more likely to form strictures after EMR only if they are nearly completely circumferential. Post-EMR strictures are usually amenable to endoscopic dilation. Preprocedure antimicrobials are generally not recommended for EMR. However, rates of bacteremia associated with submucosal injection are unknown. Current EMR techniques allow for en bloc resection for lesions ≤2 cm in diameter. Lesions >2 cm are generally removed in a piecemeal fashion. Interest has increased in endoscopic submucosal dissection for en bloc resection of larger lesions. However, this procedure has higher complication rates, demands greater procedure duration, and requires dedicated training. As the tools and techniques for endoscopic submucosal dissection develop, it may compete with EMR as the preferred technique for some lesions. Although EMR is widely considered cost effective and reduces the need for inpatient hospitalization compared with operative resection, there are no unique Current Procedural Terminology codes for this procedure. As a result, proceduralists are not adequately reimbursed for the additional time and costly specialized equipment required for this technically challenging procedure. Follow-up intervals after EMR are lesion and patient specific and based on the histopathology of the retrieved resected specimen, the endoscopist's interpretation as to the completeness of resection, lesion size, location, and need for piecemeal resection and adjunctive ablation. In most instances, surveillance endoscopy is recommended within 6–12 months after EMR. Future areas of study should consider optimal submucosal injection agents and optimal electrosurgical currents for EMR. EMR is now considered part of advanced endoscopy fellowship training, but only limited exposure to EMR is obtained in most general gastroenterology fellowships. For clinicians removed from fellowship training, EMR represents a build-on skill that may be enhanced through attending didactic and hands-on training courses. EMR has emerged as a safe and effective procedure for selected mucosal lesions throughout the gastrointestinal tract. Advanced mucosal lesions can be cured with endoscopic management, thereby obviating the need for operative resection and resulting in less morbidity and mortality with significant cost savings. Attempts at resection should be resisted unless the lesion is deemed suitable for complete eradication at a single setting and is within the scope of the procedure conferred during the informed consent process. The opportunity for referral to a practitioner of advanced EMR should be considered with the patient. If EMR is to be performed, the importance of follow-up surveillance examinations must be emphasized at the outset. Although EMR is clearly emerging as the preferred management strategy for advanced mucosal lesions, education and training are paramount to expand availability while maintaining quality.
Search from 250 M+ research papersSearch from 250 M+ research papers
Open Access
Nov 2, 2022
Simon P. Loader + 3