Advertisement

Impact of en bloc resection on long-term outcomes after endoscopic mucosal resection: a matched cohort study

Published:December 27, 2019DOI:https://doi.org/10.1016/j.gie.2019.12.025

      Background and Aims

      Residual or recurrent adenoma (RRA) is the major limitation of piecemeal EMR (p-EMR) for large colonic laterally spreading lesions (LSLs) 20 mm. En bloc EMR (e-EMR) has been shown to achieve low rates of RRA but specific procedural and long-term outcomes are unknown. Our aim was to compare long-term outcomes of size-matched LSLs stratified by whether they were resected e-EMR or p-EMR.

      Methods

      Data from a prospective tertiary referral multicenter cohort of large LSLs referred for EMR over a 10-year period were analyzed. Outcomes were compared between sized-matched LSLs (20-25 mm) resected by p-EMR or e-EMR.

      Results

      Five hundred seventy LSLs met the inclusion criteria of which 259 (45.4%) were resected by e-EMR. The risk of major deep mural injury (DMI) was significantly higher in the e-EMR group (3.5% vs 1.0%, P = .05), whereas rates of other intraprocedural adverse events did not differ significantly. Five of 9 (56%) LSLs, with endoscopic features of submucosal invasion (SMI), resected by e-EMR were saved from surgery. RRA at first surveillance was lower in the e-EMR group (2.0% vs 5.7%, P = .04), but this difference was negated at subsequent surveillance. Rates of surgical referral were not significantly different between the groups at either surveillance interval.

      Conclusion

      When comparing e-EMR against p-EMR for lesions ≤25 mm in size of similar morphology in a large prospective multicenter cohort, e-EMR offered no additional advantage for predicted-benign LSLs. However, it was associated with an increased risk of major DMI. Thus, en bloc resection techniques should be reserved for lesions suspicious for invasive disease. (Clinical trial registration number: NCT01368289.)

      Abbreviations:

      DMI (deep mural injury), e-EMR (en bloc EMR), ESD (endoscopic submucosal dissection), IQR (interquartile range), LSL (laterally spreading lesion), p-EMR (piecemeal EMR), SC1 (first surveillance colonoscopy after EMR), RRA (residual or recurrent adenoma), SC2 (second surveillance colonoscopy after EMR), SMI (submucosal invasion)

      Introduction

      EMR is accepted as the preferred endoscopic technique for the treatment of predicted-benign colonic laterally spreading lesions (LSLs) 20 mm.
      • Ferlitsch M.
      • Moss A.
      • Hassan C.
      • et al.
      Colorectal polypectomy and endoscopic mucosal resection (EMR): European Society of Gastrointestinal Endoscopy (ESGE) Clinical Guideline.
      It has been shown to be safe, efficacious, and cost-effective in comparison with surgery.
      • Moss A.
      • Williams S.J.
      • Hourigan L.F.
      • et al.
      Long-term adenoma recurrence following wide-field endoscopic mucosal resection (WF-EMR) for advanced colonic mucosal neoplasia is infrequent: results and risk factors in 1000 cases from the Australian Colonic EMR (ACE) study.
      • Jayanna M.
      • Burgess N.G.
      • Singh R.
      • et al.
      Cost analysis of endoscopic mucosal resection vs surgery for large laterally spreading colorectal lesions.
      • Ahlenstiel G.
      • Hourigan L.F.
      • Brown G.
      • et al.
      Actual endoscopic versus predicted surgical mortality for treatment of advanced mucosal neoplasia of the colon.
      Residual or recurrent adenoma (RRA) after EMR is acknowledged as an important limitation with a frequency at first surveillance of approximately 15% in a recent study.
      • Hassan C.
      • Repici A.
      • Sharma P.
      • et al.
      Efficacy and safety of endoscopic resection of large colorectal polyps: a systematic review and meta-analysis.
      Risk factors for RRA include lesion size 40 mm, high-grade dysplasia, intraprocedural bleeding requiring endoscopic control, and piecemeal resection.
      • Tate D.J.
      • Desomer L.
      • Klein A.
      • et al.
      Adenoma recurrence after piecemeal colonic EMR is predictable: the Sydney EMR recurrence tool.
      • Belderbos T.
      • Belderbos T.D.G.
      • Leenders M.
      • et al.
      Local recurrence after endoscopic mucosal resection of nonpedunculated colorectal lesions: systematic review and meta-analysis.
      • Emmanuel A.
      • Lapa C.
      • Ghosh A.
      • et al.
      Risk factors for early and late adenoma recurrence after advanced colorectal endoscopic resection at an expert Western center.
      Therefore, one approach to the problem of RRA might be to perform en bloc EMR (e-EMR) where possible. However, reliable and safe e-EMR is limited by the size of the LSL: ≤20 mm in the right colon segment and ≤25 mm in the left colon segment. Above this size, the risk of muscularis propria entrapment within the snare is high, and complete en bloc resection is unreliable.
      • Burgess N.G.
      • Bassan M.S.
      • McLeod D.
      • et al.
      Deep mural injury and perforation after colonic endoscopic mucosal resection: a new classification and analysis of risk factors.
      ,
      • Moss A.
      • Bourke M.J.
      • Tran K.
      • et al.
      Lesion isolation by circumferential submucosal incision prior to endoscopic mucosal resection (CSI-EMR) substantially improves en bloc resection rates for 40-mm colonic lesions.
      Therefore, comparing rates of RRA after piecemeal EMR (p-EMR) and e-EMR is significantly confounded by size. Yet, most previous studies have included LSLs of all sizes without description of long-term outcomes.
      To overcome this limitation, we performed a size-matched cohort analysis of the short-term and long-term outcomes of LSLs ≤25 mm removed by e-EMR compared with p-EMR.

      Methods

      Study design, inclusion and exclusion criteria

      Consecutive patients undergoing either p-EMR or e-EMR for LSLs ≥20 mm in size were enrolled over a period of 120 months (April 2009 to April 2019) at 8 tertiary referral centers as part of the Australian Colonic Endoscopic Resection Study (NCT01368289). LSLs up to 25 mm in size in the left colon segment and 20 mm in the right colon segment were eligible.
      In the case of multiple LSLs in a single patient at the same session, only the largest lesion was included in the analysis due to the potential difficulty in ascribing adverse events to a specific lesion and the risk of correlated observations in a single patient. LSLs that were not attempted or failed endoscopic resection were excluded from the analysis. Previously attempted LSLs were excluded. LSLs undergoing thermal ablation of the post-EMR margin in a separate randomized control study were also excluded (SCAR study NCT01789749).
      • Klein A.
      • Tate D.J.
      • Jayasekeran V.
      • et al.
      Thermal ablation of mucosal defect margins reduces adenoma recurrence after colonic endoscopic mucosal resection.
      Written informed consent was obtained from all participants at the index procedure. Institutional review board approval was obtained at each institution.

      Lesion classification

      Lesions were evaluated by the individual endoscopist performing EMR using high-definition white light and narrow-band imaging. Lesions were classified by the Paris classification,
      The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon: November 30 to December 1, 2002.
      morphology,
      • Tanaka S.
      • Haruma K.
      • Oka S.
      • et al.
      Clinicopathologic features and endoscopic treatment of superficially spreading colorectal neoplasms larger than 20 mm.
      and Kudo pit pattern I to V.
      • Kudo S.
      • Hirota S.
      • Nakajima T.
      • et al.
      Colorectal tumours and pit pattern.
      Lesions were assessed endoscopically for evidence of submucosal invasion (SMI) and classified as high risk if they demonstrated a demarcated area, where a regular neoplastic pit/vascular surface pattern (Kudo III/IV) appeared disordered (Kudo V). Lesion size was approximated relative to an open snare of known dimensions. Lesions of the right colon segment were defined as any lesion proximal to and including the hepatic flexure.

      Index EMR procedure

      All EMR procedures were performed by senior endoscopists with extensive experience of EMR or by an advanced endoscopy fellow under their direct supervision. Split-dose bowel preparation was used. Intravenous sedation was provided with a combination of fentanyl, midazolam, and propofol. Insufflation of the colon was initially with air moving to carbon dioxide in August 2010 once the benefits were understood.
      • Bassan M.S.
      • Holt B.
      • Moss A.
      • et al.
      Carbon dioxide insufflation reduces number of postprocedure admissions after endoscopic resection of large colonic lesions: a prospective cohort study.
      Colonoscopy was performed using Olympus 180 or 190 series high-definition variable-stiffness colonoscopes (180/190 PCF/CF; Olympus, Tokyo, Japan). A microprocessor-controlled electrosurgical generator (Endocut Q Effect 3, VIO 300D; ERBE Elektromedizin, Tübingen, Germany) with fractionated current was used in most cases. Two centers initially used ConMed 7500 electrosurgical units (ConMed, Utica, NY, USA) before transitioning to Erbe VIO 300 D in July and November 2010. Subsequently, all but one center used this generator. The remaining center used the ESG 100 system (Olympus) on Pulse Cut Slow 120W setting. The submucosal injectate comprised normal saline solution until 2010 when it was replaced with succinylated gelatin (Gelofusine; B. Braun Australia Pty Ltd, Bella Vista, Australia).
      • Moss A.
      • Field A.
      • Bourke M.J.
      • et al.
      A randomized, double-blind trial of succinylated gelatin submucosal injection for endoscopic resection of large sessile polyps of the colon.
      The colloid solution was combined with 0.4% indigo carmine blue and epinephrine diluted to 1:100,000. Occasionally, methylene blue was used as an alternative when indigo carmine was not available.
      The decision to proceed with e-EMR or p-EMR was at the discretion of the individual endoscopist. e-EMR (Fig. 1) and p-EMR (Fig. 2) was performed using the standard EMR technique as has been described previously in detail.
      • Holt B.A.
      • Bourke M.J.
      Wide field endoscopic resection for advanced colonic mucosal neoplasia: current status and future directions.
      • Klein A.
      • Bourke M.J.
      Advanced polypectomy and resection techniques.
      • Tutticci N.
      • Bourke M.J.
      Advanced endoscopic resection in the colon: recent innovations, current limitations and future directions.
      • Bourke M.
      • Jideh B.
      How to perform wide-field endoscopic mucosal resection and follow-up examinations.
      e-EMR was defined as removal of all macroscopically visible adenomatous tissue with a single snare excision. If further snare excisions were required, this was classified as p-EMR. Technical success was defined at the completion of the endoscopic procedure as complete removal of all macroscopically visible polyp tissue. Submucosal fibrosis was assessed endoscopically and graded as absent, moderate, or severe.
      Figure thumbnail gr1
      Figure 1Examples of e-EMR. A, 20 mm, Kudo III, granular laterally spreading lesions (LSLs) in the sigmoid colon. B, 25 mm, Kudo III, granular LSLs in the ascending colon. C, 20 mm, Kudo V, nongranular LSLs in the ascending colon.
      Figure thumbnail gr2
      Figure 2Examples of p-EMR. A, 25 mm, Kudo IV, granular laterally spreading lesions (LSLs) in the sigmoid colon. B, 20 mm, Kudo III, granular LSLs in the transverse colon.

      Histopathology assessment

      All e-EMR cases were pinned before histopathologic assessment and assessed for margin involvement of the resection specimen. The pathologist analyzing the specimen was blinded to the endoscopic opinion of the predicted histology.

      Study outcomes

      The primary outcome of the study was to compare the rates of surgery at the index procedure and during long-term surveillance for sized-matched lesions resected by e-EMR compared with p-EMR. Secondary endpoints included rates of adenoma recurrence during surveillance, the rate of adverse events, and procedural characteristics between the 2 techniques.

      Adverse events

      Intraprocedural bleeding was defined as that persisting for ≥30 seconds and requiring endoscopic control, primarily treated with snare-tip soft coagulation (Soft Coagulation, 80 W Effect 4; ERBE Electromedizin, Tübingen, Germany).
      • Fahrtash-Bahin F.
      • Holt B.A.
      • Jayasekeran V.
      • et al.
      Snare tip soft coagulation achieves effective and safe endoscopic hemostasis during wide-field endoscopic resection of large colonic lesions (with videos).
      Major deep mural injury (DMI) was assessed retrospectively based on prospectively collected data, as per the Sydney DMI classification,
      • Burgess N.G.
      • Bassan M.S.
      • McLeod D.
      • et al.
      Deep mural injury and perforation after colonic endoscopic mucosal resection: a new classification and analysis of risk factors.
      with a visible target sign or actual hole with or without peritoneal contamination corresponding to DMI type III to V.
      All patients were contacted by a study nurse 2 weeks after the procedure to assess for any adverse events related to their EMR procedure. Clinically significant postendoscopic bleeding was defined as bleeding after the completion of EMR and discharge from the endoscopy unit resulting in presentation to the emergency department, hospitalization, or reintervention within 14 days.
      • Burgess N.G.
      • Williams S.J.
      • Hourigan L.F.
      • et al.
      A management algorithm based on delayed bleeding after wide-field endoscopic mucosal resection of large colonic lesions.
      Delayed perforation was defined as the clinical syndrome of pain after the EMR procedure with imaging or surgical evidence of full-thickness injury to the colonic wall.

      Surveillance procedures

      Follow-up data were collected from patients eligible for first surveillance colonoscopy (SC1) at a planned interval of 4 to 6 months. Time to longest follow-up and any associated recurrence after SC1 were recorded if available. Second surveillance colonoscopy (SC2) was performed at a planned interval of 18 months after the original EMR. Time to latest follow-up was recorded in months. Late recurrence was defined as any recurrence occurring at SC2 or later at an endoscopic resection scar that was negative for recurrence at previous surveillance procedures.

      EMR scar assessment

      Recurrence was defined as the endoscopic appearance of RRA at an EMR scar unless otherwise stated. A standardized imaging protocol was used to assess the post-EMR scar for recurrence.
      • Desomer L.
      • Tutticci N.
      • Tate D.J.
      • et al.
      A standardized imaging protocol is accurate in detecting recurrence after endoscopic mucosal resection.
      To determine the presence of recurrence, a transition point was sought where a non-neoplastic pit or vascular pattern (Kudo I or II) became a neoplastic pit pattern (Kudo III or IV). Any suspected recurrence was sampled and then treated endoscopically.
      • Tate D.J.
      • Desomer L.
      • Awadie H.
      • et al.
      Simple endoscopic treatment of adenoma recurrence after wide field endoscopic mucosal resection is effective: a prospective study of 1558 lesions with long term follow up [abstract].

      Statistical analysis

      All data were analyzed using SPSS Statistics version 22.0 (IBM, Armonk, NY, USA). Categorical variables were described using frequencies and percentages. Mean, median, and interquartile ranges (IQRs) were calculated for continuous data. Statistical significance was set at a threshold of .05, and comparisons between different groups and outcomes were performed using the chi-squared test, Fisher exact test, and binary logistic regression.

      Results

      Demographics

      Over a 120-month period (April 2009 to April 2019), 3966 lesions were referred for EMR, of which 722 were eligible based on the size criterion in this study (Fig. 3). One hundred fifty-two (21.1%) LSLs were then subsequently excluded due to multiple LSLs in the same patient at the same session; EMR was not attempted due to suspected deep submucosal invasive cancer in 177, technical reasons (ileal infiltration of ileocecal valve) in 23, and deep appendiceal invasion or difficult access in 12.
      Figure thumbnail gr3
      Figure 3Study design: recruitment and follow-up. LSL, Laterally spreading lesion; SC1, first surveillance colonoscopy; SC2, second surveillance colonoscopy.
      A total of 570 lesions were included in the final analysis; the mean age of the patients was 66.7 years (53.2% male). Two hundred fifty-nine (45.4%) LSLs were resected by e-EMR and 311 (54.6%) by p-EMR. The median lesion size was 20 mm (IQR, 20-20 mm) in the e-EMR group and 20 mm (IQR, 20-25 mm) in the p-EMR group. Two hundred twenty-seven (39.8%) lesions were in the right colon segment. Submucosal fibrosis was more common in the p-EMR group than in the e-EMR group (10 [3.9%] vs 41 [13.2%], P < .001). Lesion characteristics were otherwise statistically similar between the 2 groups (Table 1).
      Table 1Characteristics of laterally spreading lesions stratified by whether they were resected e-EMR or p-EMR
      e-EMR (n = 259)p-EMR (n = 311)Total (N = 570)P valueHR (95% CI)P value
      Patients
      Age (years), mean (SD)65.45 (13.0)67.61 (11.0)66.74 (11.9).08
      Sex
       Male (%)135 (52.1)168 (54.0)303 (53.2).51
       Female (%)124 (47.9)143 (46.0)267 (46.8)
      Lesions
      Size (mm), median (IQR)20 (20-20)20 (20-25)20 (20-23).002
      Morphology (%)
       Granular99 (38.2)132 (42.4)231 (40.5).17
       Nongranular100 (38.6)120 (38.6)220 (38.6)
       Mixed36 (13.9)25 (8.0)61 (10.7)
      Kudo pit pattern (%)
      Kudo pit pattern assessed endoscopically using high-definition white light and narrow-band imaging.
       II55 (21.2)66 (21.2)121 (21.2).01
       III74 (28.6)78 (25.1)152 (26.7)
       IV88 (34.0)108 (34.7)196 (34.4)
       V (i/n)18 (6.9)8 (2.6)26 (4.6).02
      Paris classification (%)
       Is78 (30.1)72 (23.2)150 (26.3).01
       IIa146 (56.4)182 (58.5)328 (57.5)
       IIa + Is8 (3.1)28 (9.0)36 (6.3)
       IIa/IIa + c13 (5.0)9 (2.9)22 (3.8)
      Right colon109 (42.1)118 (37.9)227 (39.8).31
      Submucosal fibrosis (%)10 (3.9)41 (13.2)51<.0010.26 (0.13-0.54)<.001
      Histopathology (%)
       TA78 (30.1)111 (35.7)189 (33.2).44
       TVA104 (40.2)96 (30.9)200 (35.1)
       SSA65 (25.1)83 (26.7)148 (30.0)
      Procedure
      Successful EMR, n259296555
      Duration (minutes), median (IQR)5 (5-10)10 (10-20)10 (5-15)<.001
      IPB (%)17 (6.6)24 (8.1)41 (7.4).49
      Deep mural injury (III-V) (%)
      Sydney deep mural injury classification.9
      9 (3.5)3 (1.0)12 (2.2).053.52 (0.94-13.13).06
      Histologically clear of margins, n (%)232/259 (89.6)
      Margin positive, n (%)27/259 (10.4)
      Highest grade of dysplasia (%)
       None59 (22.8)70 (23.6)129 (23.2).83
       Low grade152 (58.7)181 (61.1)333 (60.0)
       High grade48 (18.5)45 (15.2)93 (16.8)
      Invasive cancer (%)15 (5.8)15 (5.1)30 (5.4).70
      Adverse events
       CSPEB (%)9 (3.5)8 (2.7)17 (3.1).60
       Delayed perforation (%)1 (0.4)0 (0)1 (1.8).29
       Surgery after technically successful index procedure, n (%)15 (5.8)11 (3.7)26 (4.7).251.59 (0.72-3.53).25
      e-EMR, En bloc EMR; p-EMR, piecemeal EMR; HR, hazard ratio; CI, confidence interval; SD, standard deviation; IQR, interquartile range; TA, tubular adenoma; TVA, tubulo-villous adenoma; SSA, sessile serrated adenoma; IPB, intraprocedural bleeding; CSPEB, clinically significant postendoscopic bleeding.
      Kudo pit pattern assessed endoscopically using high-definition white light and narrow-band imaging.
      Sydney deep mural injury classification.
      • Burgess N.G.
      • Bassan M.S.
      • McLeod D.
      • et al.
      Deep mural injury and perforation after colonic endoscopic mucosal resection: a new classification and analysis of risk factors.

      Procedure

      Median procedure duration was significantly shorter in the e-EMR group than with p-EMR (5 minutes [IQR, 5-10 minutes] vs 10 minutes [IQR, 10-20 minutes], P < .001). Technical success at EMR was achieved in all 259 lesions in the e-EMR cohort and 296 of 311 (95.2%) in the p-EMR group (P = .09). In total, 98 (17.2%) cases in this cohort were resected using normal saline solution in the submucosal injectate, and the remainder were resected using succinylated gelatin.

      Intraprocedural adverse events

      Intraprocedural bleeding was not significantly different between the p-EMR and e-EMR groups (17 [6.6%] vs 24 [8.1%], P = .36). Major DMI was significantly more common in the LSLs removed by e-EMR (9 [3.5%] vs 3 [1.0%], P = .05; hazard ratio [HR], 3.53; 95% confidence interval, 0.9-13.13). At the subgroup analysis, there was no significant difference in the rate of major DMI in lesions resected using normal saline solution (3 of 63 [4.8%] e-EMR vs 0 of 29 [0%] p-EMR, P = .23) in the submucosal injectate compared with succinylated gelatin (7 of 196 [3.6%] e-EMR vs 9 of 282 [3.2%] p-EMR, P = .82).

      Post-EMR adverse events

      There was no significant difference in the rate of clinically significant postendoscopic bleeding between the 2 groups (9 [3.5%] vs 8 [2.7%], P = .60). The single case of delayed perforation occurred after en bloc resection of a 20-mm granular LSL in the left colon segment.

      Margins

      Of the lesions successfully resected in e-EMR group, 89.6% (232 of 259) had negative margins at histopathologic assessment.

      LSLs with endoscopic suspicion of SMI

      Eighteen (6.9%) LSLs in the e-EMR group and 8 (2.6%) in the p-EMR group had endoscopic imaging consistent with high risk for SMI. Of the 18 high-risk LSLs in the e-EMR group, 9 had evidence of invasive cancer at histopathology, of which 5 (56%) were assessed as low risk for SMI and considered endoscopically cured.
      • Pimentel-Nunes P.
      • Dinis-Ribeiro M.
      • Ponchon T.
      • et al.
      Endoscopic submucosal dissection: European Society of Gastrointestinal Endoscopy (ESGE) Guideline.
      Of the 8 high-risk LSLs in p-EMR group, 4 (50%) had evidence of invasive cancer at histopathology but could not be considered for endoscopic cure due to piecemeal resection.

      Surgery after a successful index procedure

      Fifteen of 259 (5.8%) LSLs in the e-EMR group and 11 of 296 (3.7%) LSLs in the p-EMR group underwent surgery (P = .25). Eight lesions with invasive cancer were not suited for surgery due to patient comorbidity. Twenty-two of 26 (84.6%) cases underwent surgery due to invasive cancer (Supplementary Table 1, available online at www.giejournal.org). The remainder underwent surgery either due to delayed perforation or for a reason unrelated to the EMR procedure. Ten of 22 (45.5%) lesions that underwent surgery due to unfavorable histopathology did not have residual cancer in the surgical specimen.

      Recurrence

      Two hundred one patients in the e-EMR group (90.1%) and 247 patients in the p-EMR group (93.6%) eligible for follow-up underwent their first surveillance colonoscopy at a median 5.9 months (IQR, 4-9.7 months) (Table 2). Recurrence at SC1 was significantly less common among LSLs resected by e-EMR compared with p-EMR (4 [2%] vs 14 [5.7%], P = .04, HR, 3.02 [95% confidence interval, 0.98-9.32], P = .06). Of the 4 patients with recurrence in the e-EMR group, all had negative histologic margins at the index EMR. Three patients were referred for surgery after SC1 (2 in the e-EMR group, 1 in the p-EMR group). All 3 cases were referred due to lesion recurrence that could not be treated endoscopically. Statistically, there was no difference in rates of surgery at SC1 between the groups (2 [1%] vs 1 [0.4%], P = .47). One hundred forty-eight (75.5%) eligible patients in the e-EMR group and 164 (72.6%) eligible patients in the p-EMR group underwent SC2 at a median 19 months (IQR, 16-19 months). Submucosal injectate composition did not influence the recurrence rate at SC1; normal saline solution (2 of 53 [3.8%] e-EMR vs 2 of 24 [4.2%] p-EMR, P = .93), succinylated gelatin (2 of 152 [1.3%] e-EMR vs 15 of 233 [6.4%] p-EMR, P = .10).
      Table 2Outcomes at follow-up of laterally spreading lesions after a technically successful index EMR stratified by whether they were e-EMR or p-EMR∗
      e-EMRp-EMRTotalP valueOR (95% CI)P value
      First surveillance colonoscopy (SC1), n201247448
      Months, median (IQR)6.7 (4.1-12.0)5.3 (4.0-8.0)5.9 (4.0-9.7).001
      Recurrence, n (%)4 (2.0)14 (5.7)18.043.02 (0.98-9.32).06
      Surgery after SC1, n (%)2 (1.0)1 (0.4)3.47
      Second surveillance colonoscopy (SC2), n148164312
      Months, median (IQR)22.0 (16.0-30.0)18.0 (15-22.0)19.0 (16.0-19.0).01
      Recurrence, n (%)2 (1.4)4 (2.4)6.49
      Surgery after SC2, n (%)1 (0.7)0 (0)0-
      Late recurrence, n (%)0 (0)1 (1.1)1.40
      Data based on LSLs eligible for follow-up (see Fig. 3).
      e-EMR, En bloc EMR; p-EMR, piecemeal EMR; OR, odds ratio; CI, confidence interval; IQR, interquartile range.
      *Percentage calculated from those who underwent SC1/SC2.
      Recurrence at SC2 was not significantly different between the groups (2 [1.4%] vs 4 [2.4%], P = .49). There was also no difference in the rate of surgical referral at SC2; only 1 case in the e-EMR group was referred due to an inability to treat recurrence endoscopically. The median time to last surveillance colonoscopy was 39.50 months (IQR, 28.25-53.75 months).

      Discussion

      e-EMR, compared with p-EMR, has been shown to reduce the rate of RRA for LSLs 20 mm.
      • Moss A.
      • Williams S.J.
      • Hourigan L.F.
      • et al.
      Long-term adenoma recurrence following wide-field endoscopic mucosal resection (WF-EMR) for advanced colonic mucosal neoplasia is infrequent: results and risk factors in 1000 cases from the Australian Colonic EMR (ACE) study.
      ,
      • Tate D.J.
      • Desomer L.
      • Klein A.
      • et al.
      Adenoma recurrence after piecemeal colonic EMR is predictable: the Sydney EMR recurrence tool.
      ,
      • Belderbos T.
      • Belderbos T.D.G.
      • Leenders M.
      • et al.
      Local recurrence after endoscopic mucosal resection of nonpedunculated colorectal lesions: systematic review and meta-analysis.
      However, to date, large-volume size-matched series with long-term follow-up have not been described. In this study, e-EMR was compared with p-EMR for LSLs up to 25 mm. These lesions represent the recognized upper limit of reliable and safe en bloc resection and are at low risk of submucosal invasion.
      • Ferlitsch M.
      • Moss A.
      • Hassan C.
      • et al.
      Colorectal polypectomy and endoscopic mucosal resection (EMR): European Society of Gastrointestinal Endoscopy (ESGE) Clinical Guideline.
      ,
      • Burgess N.G.
      • Hourigan L.F.
      • Zanati S.A.
      • et al.
      Risk stratification for covert invasive cancer among patients referred for colonic endoscopic mucosal resection: a large multicenter cohort.
      This study demonstrates that although e-EMR should be considered for the small subset of these lesions at high risk for SMI, it is associated with an increased risk of intraprocedural adverse events with no clear long-term benefit in rates of RRA or need for surgery over p-EMR.
      One of the major advantages of e-EMR is the superior histologic specimen obtained. A single histologic specimen demonstrates completeness of LSL resection
      • Sobin L.H.
      • Fleming I.D.
      TNM Classification of Malignant Tumors, fifth edition (1997). Union Internationale Contre le Cancer and the American Joint Committee on Cancer.
      and the presence or absence of invasion into the submucosa. For colorectal neoplasia confined to the mucosa, this may be of limited benefit due to the absence of colonic mucosal lymphatics and consequent improbability of distant neoplastic spread. Complete resection may be alternatively assessed with a detailed endoscopic examination of the post-EMR defect
      • Desomer L.
      • Tate D.J.
      • Bahin F.F.
      • et al.
      A systematic description of the post-EMR defect to identify risk factors for clinically significant post-EMR bleeding in the colon.
      and confirmed by a thorough assessment of the endoscopic resection scar during surveillance examinations.
      • Desomer L.
      • Tutticci N.
      • Tate D.J.
      • et al.
      A standardized imaging protocol is accurate in detecting recurrence after endoscopic mucosal resection.
      ,
      • Kandel P.
      • Brand E.C.
      • Pelt J.
      • et al.
      Endoscopic scar assessment after colorectal endoscopic mucosal resection scars: when is biopsy necessary (EMR Scar Assessment Project for Endoscope (ESCAPE) trial).
      However, for lesions with SMI, an en bloc specimen is of the utmost importance, because demonstration of the muscularis mucosa in its entirety is desirable to assess the width and depth of any invasive focus and the presence of high-risk features such as lymphovascular invasion or tumor budding. In accordance with current guidelines, lesions with submucosal infiltration <1000 μm and lacking the aforementioned high-risk features can be cured endoscopically if en bloc resection is achieved, due to the low rate of lymph node metastasis.
      • Pimentel-Nunes P.
      • Dinis-Ribeiro M.
      • Ponchon T.
      • et al.
      Endoscopic submucosal dissection: European Society of Gastrointestinal Endoscopy (ESGE) Guideline.
      ,
      • Kikuchi R.
      • Takano M.
      • Takagi K.
      • et al.
      Management of early invasive colorectal cancer. Risk of recurrence and clinical guidelines.
      ,
      • Beaton C.
      • Twine C.P.
      • Williams G.L.
      • et al.
      Systematic review and meta-analysis of histopathological factors influencing the risk of lymph node metastasis in early colorectal cancer.
      The strongest suspicion of SMI in a colonic LSL is raised by the detection of a demarcated area, often depressed, where a regular neoplastic pit/vascular pattern (Kudo III/IV) becomes disordered (Kudo V).
      • Kudo S.
      • Hirota S.
      • Nakajima T.
      • et al.
      Colorectal tumours and pit pattern.
      ,
      • Hayashi N.
      • Tanaka S.
      • Hewett D.G.
      • et al.
      Endoscopic prediction of deep submucosal invasive carcinoma: validation of the Narrow-Band Imaging International Colorectal Endoscopic (NICE) classification.
      In this study, a benefit of resecting LSLs with such overt features, confirmed as invasive cancer at histology, was shown with 5 of 9 (56%) such patients considered endoscopically cured (due to the absence of high-risk histologic features and R0 resection). Endoscopists could therefore strongly consider en bloc resection in the rare cases in which overt features of SMI are encountered.
      A further minority of LSLs may harbor a small focus of SMI and, when lacking the typical surface features mentioned above, evade endoscopic detection. In a recent study by Burgess et al,
      • Burgess N.G.
      • Hourigan L.F.
      • Zanati S.A.
      • et al.
      Risk stratification for covert invasive cancer among patients referred for colonic endoscopic mucosal resection: a large multicenter cohort.
      the risk of SMI in such lesions can be stratified based on lesion morphology, location, and increasing lesion size. For lesions 25 mm or less in size, the risk is minimal, and in this study, the cohort comprised <3% of these lesions.
      The findings from this study confirm that the rate of RRA at first surveillance is significantly higher when size-matched LSLs are removed p-EMR versus e-EMR. However, the current data demonstrate that this difference is negated at the second surveillance procedure (median, 19 months). Moreover the primary endpoint of avoidance of surgery was nonsignificant between LSLs resected by p-EMR or e-EMR at both surveillance time points, primarily due to the ability to treat recurrence endoscopically.
      • Tate D.J.
      • Desomer L.
      • Awadie H.
      • et al.
      Simple endoscopic treatment of adenoma recurrence after wide field endoscopic mucosal resection is effective: a prospective study of 1558 lesions with long term follow up [abstract].
      Taken together, these findings suggest that e-EMR and p-EMR are equivalent in terms of important long-term outcomes for LSLs at low risk of SMI.
      A recently published randomized study
      • Klein A.
      • Tate D.J.
      • Jayasekeran V.
      • et al.
      Thermal ablation of mucosal defect margins reduces adenoma recurrence after colonic endoscopic mucosal resection.
      demonstrated that thermal ablation of the post-EMR margin led to a 4-fold reduction in the rate of RRA (21% vs 5.2%). A post-hoc analysis of that study demonstrated that the effect of e-EMR was negated by the effect of the active treatment. Lesions resected by e-EMR in the control arm had a nonsignificant difference in the rate of RRA compared with those undergoing p-EMR in the active arm. This highly effective treatment to prevent RRA was not used in this study. This further emphasizes the advantage of the modern p-EMR technique, including complete thermal ablation of the post-EMR margin, over an e-EMR strategy.
      The results of the current study also demonstrate potential procedural reasons why endoscopists may decide not to pursue en bloc resection for LSLs at low risk for SMI. The risk of major DMI (type III-V) was significantly increased when similar-sized LSLs were removed by e-EMR versus p-EMR. These results are consistent with outcomes from previously published studies assessing intraprocedural adverse events of e-EMR.
      • Burgess N.G.
      • Bassan M.S.
      • McLeod D.
      • et al.
      Deep mural injury and perforation after colonic endoscopic mucosal resection: a new classification and analysis of risk factors.
      We also observed a nonsignificant trend toward an increased rate of delayed bleeding after e-EMR. Taken together, these observations suggest that for LSLs up to 25 mm in size, the risks to the patient are higher when using e-EMR. Major DMI, if recognized, can be managed by endoscopic clip closure without sequelae. If signs of major DMI are not recognized, then delayed perforation is a risk, often resulting in a poor clinical outcome. Moreover, the single delayed perforation in this cohort occurred after e-EMR of a 20-mm LSL in the left colon segment.
      Furthermore, we observed in this study that, even in a tertiary endoscopy setting where the endoscopist intends to perform e-EMR, more than 10% of macroscopically complete resections do not achieve R0 resection histologically. This raises the possibility that, while performing e-EMR, the endoscopist may inadvertently leave residual microscopic adenoma at the defect margin. To mitigate this high rate of positive histologic margins in e-EMR, perhaps thermal ablation of the defect margin could be performed routinely. Conversely, of the 4 cases of recurrence in the e-EMR group at SC1, all lesions were histologically clear of the resection margins at the index EMR when assessed by specialist GI pathologists. This therefore highlights the importance of meticulous identification and inspection of all EMR scars at surveillance intervals, even for cases of R0 resection.
      We recognize other techniques are available that may allow for more reliable rates of en bloc resection and for inclusion of larger LSLs. These include underwater EMR,
      • Yamashina T.
      • Uedo N.
      • Akasaka T.
      • et al.
      Comparison of underwater vs conventional endoscopic mucosal resection of intermediate-size colorectal polyps.
      full-thickness endoscopic resection,
      • Kuellmer A.
      • Mueller J.
      • Caca K.
      • et al.
      Endoscopic full-thickness resection for early colorectal cancer.
      and endoscopic submucosal dissection (ESD).
      • Saito Y.
      • Uraoka T.
      • Yamaguchi Y.
      • et al.
      A prospective, multicenter study of 1111 colorectal endoscopic submucosal dissections (with video).
      However, compared with e-EMR, these techniques are associated with significantly longer procedure duration, increased resource utilization, higher rates of intraprocedural adverse events, and need for extensive technical training.
      • Yamashina T.
      • Uedo N.
      • Akasaka T.
      • et al.
      Comparison of underwater vs conventional endoscopic mucosal resection of intermediate-size colorectal polyps.
      ,
      • Spadaccini M.
      • Fuccio L.
      • Lamonaca L.
      • et al.
      Underwater EMR for colorectal lesions: a systematic review with meta-analysis (with video).
      • Terasaki M.
      • Tanaka S.
      • Oka S.
      • et al.
      Clinical outcomes of endoscopic submucosal dissection and endoscopic mucosal resection for laterally spreading tumors larger than 20 mm.
      • Bahin F.F.
      • Heitman S.J.
      • Bourke M.J.
      Wide-field endoscopic mucosal resection versus endoscopic submucosal dissection for laterally spreading colorectal lesions: a cost-effectiveness analysis.
      In addition, data supporting their routine use for benign LSLs of 20 to 25 mm are less robust.
      The strengths of this study include the inclusion of consecutive patients enrolled prospectively across 8 tertiary referral endoscopy centers and the high compliance to long-term surveillance. We appreciate the limitations of our study. Our cohort represents a small subset of lesions referred for tertiary EMR, and therefore our findings cannot be generalized to lesions outside the size-based lesion inclusion criteria. Our dataset did not allow us to identify lesions that were initially intended for e-EMR but were subsequently treated by p-EMR for technical reasons. This may have led to biased inclusion in the p-EMR group. Endoscopists in this study also did not have access to ESD, which may have led them to attempt en bloc resection for lesions better suited to other resection modalities such as ESD. However, by limiting the lesion size in the study, this potential bias has been limited.

      Conclusion

      In summary, in a large multicenter cohort of LSLs referred for EMR, e-EMR and p-EMR are equally efficacious for the treatment of colonic LSLs ≤25 mm. There was no significant difference in the rates of adenoma recurrence or surgery in long-term follow-up. In addition, due to the higher rates of intraprocedural adverse events observed, e-EMR may unnecessarily expose patients to an increase risk of adverse clinical outcomes without a meaningful clinical benefit. Moreover, these data predate the use of thermal ablation of the post-EMR defect margin which, if implemented routinely, is likely to further reduce the risk of recurrence and the perceived benefit of e-EMR for 20- to 25-mm LSLs. Thus, LSLs ≤25 mm can be safely removed by p-EMR with e-EMR or alternate en bloc resection techniques reserved for lesions suspicious for invasive disease.

      Acknowledgments

      The Cancer Institute New South Wales provided funding for a research nurse to assist with the administration of the study. There was no influence from the institution regarding study design or conduct, data collection, management, analysis or interpretation, or preparation, review, or approval of the manuscript.

      Appendix

      Supplementary Table 1Laterally spreading lesions that underwent surgery after technically successful index EMR with subsequent histopathology
      IndexMorphologyLocationReason for surgeryResidual cancer
      e-EMR
      Lesion 1NGLeftInvasive diseaseNo residual cancer
      Lesion 2NGLeftInvasive diseaseT2N1
      Lesion 3GLeftInvasive diseaseNo residual cancer
      Lesion 4NGLeftColon surgery not related to EMR
      Lesion 5GLeftDelayed perforation
      Lesion 6GLeftInvasive diseaseNo residual cancer
      Lesion 7NGLeftInvasive diseaseT1
      Lesion 8NGLeftInvasive diseaseT2N1
      Lesion 9SerratedLeftInvasive diseaseNo residual cancer
      Lesion 10SerratedLeftInvasive diseaseNo residual cancer
      Lesion 11GLeftInvasive diseaseT2
      Lesion 12GRightInvasive diseaseT1
      Lesion 13NGLeftInvasive diseaseT2
      Lesion 14NGLeftInvasive diseaseT1N0
      Lesion 15MixedLeftInvasive diseaseT2
      p-EMR
      Lesion 1NGLeftInvasive diseaseT2N1
      Lesion 2NGLeftInvasive diseaseNo residual cancer
      Lesion 3NGLeftInvasive diseaseNo residual cancer
      Lesion 4MixedLeftInvasive diseaseT3N1
      Lesion 5NGLeftColon surgery not related to EMR
      Lesion 6MixedLeftInvasive diseaseNo residual cancer
      Lesion 7NGRightInvasive diseaseNo residual cancer
      Lesion 8SerratedLeftColon surgery not related to EMR
      Lesion 9GRightInvasive diseaseNo residual cancer
      Lesion 10SSALeftInvasive diseaseT3N1
      Lesion 11N/ALeftInvasive diseaseT2
      NG, Nongranular; G, granular; SSA, sessile serrated adenoma.

      References

        • Ferlitsch M.
        • Moss A.
        • Hassan C.
        • et al.
        Colorectal polypectomy and endoscopic mucosal resection (EMR): European Society of Gastrointestinal Endoscopy (ESGE) Clinical Guideline.
        Endoscopy. 2017; 49: 270-297
        • Moss A.
        • Williams S.J.
        • Hourigan L.F.
        • et al.
        Long-term adenoma recurrence following wide-field endoscopic mucosal resection (WF-EMR) for advanced colonic mucosal neoplasia is infrequent: results and risk factors in 1000 cases from the Australian Colonic EMR (ACE) study.
        Gut. 2015; 64: 57-65
        • Jayanna M.
        • Burgess N.G.
        • Singh R.
        • et al.
        Cost analysis of endoscopic mucosal resection vs surgery for large laterally spreading colorectal lesions.
        Clin Gastroenterol Hepatol. 2016; 14: 271-278.e1-2
        • Ahlenstiel G.
        • Hourigan L.F.
        • Brown G.
        • et al.
        Actual endoscopic versus predicted surgical mortality for treatment of advanced mucosal neoplasia of the colon.
        Gastrointest Endosc. 2014; 80: 668-676
        • Hassan C.
        • Repici A.
        • Sharma P.
        • et al.
        Efficacy and safety of endoscopic resection of large colorectal polyps: a systematic review and meta-analysis.
        Gut. 2016; 65: 806-820
        • Tate D.J.
        • Desomer L.
        • Klein A.
        • et al.
        Adenoma recurrence after piecemeal colonic EMR is predictable: the Sydney EMR recurrence tool.
        Gastrointest Endosc. 2017; 85: 647-656.e6
        • Belderbos T.
        • Belderbos T.D.G.
        • Leenders M.
        • et al.
        Local recurrence after endoscopic mucosal resection of nonpedunculated colorectal lesions: systematic review and meta-analysis.
        Endoscopy. 2014; 46: 388-402
        • Emmanuel A.
        • Lapa C.
        • Ghosh A.
        • et al.
        Risk factors for early and late adenoma recurrence after advanced colorectal endoscopic resection at an expert Western center.
        Gastrointest Endosc. 2019; 90: 127-136
        • Burgess N.G.
        • Bassan M.S.
        • McLeod D.
        • et al.
        Deep mural injury and perforation after colonic endoscopic mucosal resection: a new classification and analysis of risk factors.
        Gut. 2016; 66: 1779-1789
        • Moss A.
        • Bourke M.J.
        • Tran K.
        • et al.
        Lesion isolation by circumferential submucosal incision prior to endoscopic mucosal resection (CSI-EMR) substantially improves en bloc resection rates for 40-mm colonic lesions.
        Endoscopy. 2010; 42: 400-404
        • Klein A.
        • Tate D.J.
        • Jayasekeran V.
        • et al.
        Thermal ablation of mucosal defect margins reduces adenoma recurrence after colonic endoscopic mucosal resection.
        Gastroenterology. 2018; 156: 604-613.e3
      1. The Paris endoscopic classification of superficial neoplastic lesions: esophagus, stomach, and colon: November 30 to December 1, 2002.
        Gastrointest Endosc. 2003; 58: S3-S43
        • Tanaka S.
        • Haruma K.
        • Oka S.
        • et al.
        Clinicopathologic features and endoscopic treatment of superficially spreading colorectal neoplasms larger than 20 mm.
        Gastrointest Endosc. 2001; 54: 62-66
        • Kudo S.
        • Hirota S.
        • Nakajima T.
        • et al.
        Colorectal tumours and pit pattern.
        J Clin Pathol. 1994; 47: 880-885
        • Bassan M.S.
        • Holt B.
        • Moss A.
        • et al.
        Carbon dioxide insufflation reduces number of postprocedure admissions after endoscopic resection of large colonic lesions: a prospective cohort study.
        Gastrointest Endosc. 2013; 77: 90-95
        • Moss A.
        • Field A.
        • Bourke M.J.
        • et al.
        A randomized, double-blind trial of succinylated gelatin submucosal injection for endoscopic resection of large sessile polyps of the colon.
        Am J Gastroenterol. 2010; 105: 2375-2382
        • Holt B.A.
        • Bourke M.J.
        Wide field endoscopic resection for advanced colonic mucosal neoplasia: current status and future directions.
        Clin Gastroenterol Hepatol. 2012; 10: 969-979
        • Klein A.
        • Bourke M.J.
        Advanced polypectomy and resection techniques.
        Gastrointest Endosc Clin N Am. 2015; 25: 303-333
        • Tutticci N.
        • Bourke M.J.
        Advanced endoscopic resection in the colon: recent innovations, current limitations and future directions.
        Expert Rev Gastroenterol Hepatol. 2014; 8: 161-177
        • Bourke M.
        • Jideh B.
        How to perform wide-field endoscopic mucosal resection and follow-up examinations.
        Gastrointest Endosc Clin N Am. 2019; 29: 629-646
        • Fahrtash-Bahin F.
        • Holt B.A.
        • Jayasekeran V.
        • et al.
        Snare tip soft coagulation achieves effective and safe endoscopic hemostasis during wide-field endoscopic resection of large colonic lesions (with videos).
        Gastrointest Endosc. 2013; 78: 158-163.e1
        • Burgess N.G.
        • Williams S.J.
        • Hourigan L.F.
        • et al.
        A management algorithm based on delayed bleeding after wide-field endoscopic mucosal resection of large colonic lesions.
        Clin Gastroenterol Hepatol. 2014; 12: 1525-1533
        • Desomer L.
        • Tutticci N.
        • Tate D.J.
        • et al.
        A standardized imaging protocol is accurate in detecting recurrence after endoscopic mucosal resection.
        Gastrointest Endosc. 2016; 85: 518-526
        • Tate D.J.
        • Desomer L.
        • Awadie H.
        • et al.
        Simple endoscopic treatment of adenoma recurrence after wide field endoscopic mucosal resection is effective: a prospective study of 1558 lesions with long term follow up [abstract].
        Gastrointest Endosc. 2018; 87: AB100-AB101
        • Pimentel-Nunes P.
        • Dinis-Ribeiro M.
        • Ponchon T.
        • et al.
        Endoscopic submucosal dissection: European Society of Gastrointestinal Endoscopy (ESGE) Guideline.
        Endoscopy. 2015; 47: 829-854
        • Burgess N.G.
        • Hourigan L.F.
        • Zanati S.A.
        • et al.
        Risk stratification for covert invasive cancer among patients referred for colonic endoscopic mucosal resection: a large multicenter cohort.
        Gastroenterology. 2017; 153: 732-742.e1
        • Sobin L.H.
        • Fleming I.D.
        TNM Classification of Malignant Tumors, fifth edition (1997). Union Internationale Contre le Cancer and the American Joint Committee on Cancer.
        Cancer. 1997; 80: 1803-1804
        • Desomer L.
        • Tate D.J.
        • Bahin F.F.
        • et al.
        A systematic description of the post-EMR defect to identify risk factors for clinically significant post-EMR bleeding in the colon.
        Gastrointest Endosc. 2019; 89: 614-624
        • Kandel P.
        • Brand E.C.
        • Pelt J.
        • et al.
        Endoscopic scar assessment after colorectal endoscopic mucosal resection scars: when is biopsy necessary (EMR Scar Assessment Project for Endoscope (ESCAPE) trial).
        Gut. 2019; 68: 1633-1641
        • Kikuchi R.
        • Takano M.
        • Takagi K.
        • et al.
        Management of early invasive colorectal cancer. Risk of recurrence and clinical guidelines.
        Dis Colon Rectum. 1995; 38: 1286-1295
        • Beaton C.
        • Twine C.P.
        • Williams G.L.
        • et al.
        Systematic review and meta-analysis of histopathological factors influencing the risk of lymph node metastasis in early colorectal cancer.
        Colorectal Dis. 2013; 15: 788-797
        • Hayashi N.
        • Tanaka S.
        • Hewett D.G.
        • et al.
        Endoscopic prediction of deep submucosal invasive carcinoma: validation of the Narrow-Band Imaging International Colorectal Endoscopic (NICE) classification.
        Gastrointest Endosc. 2013; 78: 625-632
        • Yamashina T.
        • Uedo N.
        • Akasaka T.
        • et al.
        Comparison of underwater vs conventional endoscopic mucosal resection of intermediate-size colorectal polyps.
        Gastroenterology. 2019; 157: 451-452
        • Kuellmer A.
        • Mueller J.
        • Caca K.
        • et al.
        Endoscopic full-thickness resection for early colorectal cancer.
        Gastrointest Endosc. 2019; 89: 1180-1181
        • Saito Y.
        • Uraoka T.
        • Yamaguchi Y.
        • et al.
        A prospective, multicenter study of 1111 colorectal endoscopic submucosal dissections (with video).
        Gastrointest Endosc. 2010; 72: 1217-1225
        • Spadaccini M.
        • Fuccio L.
        • Lamonaca L.
        • et al.
        Underwater EMR for colorectal lesions: a systematic review with meta-analysis (with video).
        Gastrointest Endosc. 2019; 89: 1109-1116.e4
        • Terasaki M.
        • Tanaka S.
        • Oka S.
        • et al.
        Clinical outcomes of endoscopic submucosal dissection and endoscopic mucosal resection for laterally spreading tumors larger than 20 mm.
        J Gastroenterol Hepatol. 2012; 27: 734-740
        • Bahin F.F.
        • Heitman S.J.
        • Bourke M.J.
        Wide-field endoscopic mucosal resection versus endoscopic submucosal dissection for laterally spreading colorectal lesions: a cost-effectiveness analysis.
        Gut. 2018; 67: 1965-1973

      Linked Article