Endoscopic Closure After Colorectal ESD: A Literature Review and Meta-Analysis on Its Efficacy in Preventing Adverse Events

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is a technical paper reviewing on existing techniques employed for defect closure after ESD. It is a comprehensive paper incorporating most existing platform.

1) Methods/review process needs to be further specified

2) Grouping of methods (clip suturing etc should be made in the text so that readers can more easily understand the information given), as was done in Table 2.

3) Discussion part could be enriched by incorporating some strategies/conclusions/decision making by the authors according to the evidence given in the aforementioned paragraphs.

4) How was table 4 constructed?was meta-analysis performed? Please provide more data and explanations

5) Line 302, authors state data taken together? Does this refer to meta-analysis?

6) Line 317, authors fail to explain how they reach those conclusions

 

The quality of English in good and does not need improving.

Author Response

Reviewer 1

This is a technical paper reviewing on existing techniques employed for defect closure after ESD. It is a comprehensive paper incorporating most existing platform.

 

1) Methods/review process needs to be further specified

 

Answer: Thank you for your valuable comments. We have clarified our methods, including the review process, as described below. The explanations for Table 2 and Tables 3/4 differed because the analyses for Tables 3/4 were conducted as a form of simplified meta-analysis. Therefore, we additionally performed a pooled analysis using forest plots and incorporated these results into the revised manuscript including abstract, supplemental table, and new Figure 6-8.

 

For Table 2

A literature search of PubMed up to April 2026 was performed using the key terms (“endoscopic closure” “suture” OR “Clip”) AND (“endoscopic submucosal dissection” OR “ESD”) AND (“Colorectum” “Colon” OR “Colorectal”) for examining the efficacy of various closure methods in this review. Studies on the efficacy of endoscopic prophy-lactic closure involving ≥5 cases were extracted. When a technique initially reported in fewer than 10 cases was later published in a study including more than 10 cases, the later report was included. The former one was remained as appropriate. However, studies whose techniques and devices were overlapped were excluded. Studies which we could not extract the data of a colorectum from the overall data of several organs were also excluded. The clinical outcomes were measured as the rates of DB, DP, and PECS. The technical outcomes were defect size, lesion size (when defect size were not available), complete closure rate, closure time, A previous comprehensive review of 13 reports on closure techniques published up to 2021 reported a high overall mean clo-sure rate of 95.7%, although procedure times varied widely from 8.0 to 56.0 minutes (mean 16.7 minutes) [7]. Including these reports, we reviewed 22 clinical studies pub-lished up to April 2026 evaluating various closure methods for colorectal ESD defects (Table 2) [20–41].

 

For Table 3/4

Paragraph 6.

Efficacy of Complete Closure for DB, DP, and PECS According to Reported Studies Including Randomized Controlled Trials and Retrospective Studies: A simplified meta-analysis

 

Method

We searched PubMed, Embase, and the Cochrane Library for eligible studies up to April 2026 for a simplified meta-analysis about the efficacy of endoscopic closure. The key terms were (“closure” “suture” OR “Clip”) AND (“endoscopic submucosal dissection” OR “ESD”) AND (“Colorectum” “Colon” OR “Colorectal”) for examining the efficacy of various closure methods in this review. In this simplified meta-analysis, we evaluated the efficacy of endoscopic closure after colorectal ESD for preventing DB, DP, and PECS. The numbers of adverse events and total cases in the closure and non-closure groups were extracted from each eligible study. The definition of complete resection and each adverse events were followed by each report’s one. Inclusion criteria were (i) patients undergoing colorectal ESD; (ii) prophylactic endoscopic closure after ESD; (iii) existence of non-closure group; (iv) incidence of either DB, DP, and PECS; (v) randomized controlled trials (RCTs) or observational studies (prospective or retrospective, case-control, or cohort studies); (vi) publication in English. Two investigators (N.Y. and Y.T.) independently extracted the data and assessed study quality. Any discrepancies were re-solved by discussion.

 

Statistical analysis

Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for each outcome. When a study included a zero-event cell, a continuity correction of 0.5 was applied. Studies with no events in both groups were excluded from pooled effect estimation for the corresponding outcome. Pooled analyses were conducted using a random-effects model according to the DerSimonian–Laird method because clinical and methodological heterogeneity among studies was anticipated. Statistical heterogeneity was assessed using Cochran’s Q test and the I² statistic. An I² value >50% was considered indicative of substantial heterogeneity. Forest plots were generated for DB, DP, and PECS, including study weights, pooled estimates, and 95% CIs. Definitions of adverse events varied slightly among included studies, and the original definitions used in each study were accepted for analysis. All statistical analyses were performed using Python-based statistical scripts. Statistical analyses and forest plots were generated using Review Manager (RevMan) version 5.4 and Python version 3.10. Statistical heterogeneity was assessed using the I² statistic and Cochran’s Q test.

 

Results

A total of 520 studies were identified through database searching. After screening and eligibility assessment, 22 studies were included in the simplified meta-analysis (Table 4) [41,44-56]. This meta-analysis was performed for DB, DP, and PECS comparing closure and non-closure groups after colorectal ESD. The methodologic quality of the 4 RCTs and 18 retrospective studies were 7-8 and 5-8 according to the Newcastle-Ottawa scale (Supple-mental Table 1) [56.1].

Regarding the simplified meta-analysis, for DB, pooled analysis demonstrated that endoscopic complete closure was significantly associated with a reduced risk of DB (OR, 0.77; 95% CI, 0.60–0.97; p=0.030; I²=53.6%) (Figure 6). For DP, there was no significant difference between the closure and non-closure groups (OR, 0.60; 95% CI, 0.23–1.55; p=0.290; I²=0.0%) (Figure 7). For PECS, endoscopic closure was not significantly associated with risk reduction (OR, 0.94; 95% CI, 0.65–1.38; p=0.765; I²=56.6%) (Figure 8).

 

Figure 6

Forest plot of the included studies evaluating the efficacy of complete closure for DB after colorectal ESD

 

Figure 7

Forest plot of the included studies evaluating the efficacy of complete closure for DP after colorectal ESD

 

Figure 8

Forest plot of the included studies evaluating the efficacy of complete closure for PECS after colorectal ESD

 

Limitations

Several limitations should be considered in this meta-analysis. The definitions of DB, DP, and PECS were not completely uniform among the included studies, which may have contributed to heterogeneity. Most included studies were retrospective observational studies, and only a limited number of randomized controlled trials were available. Closure techniques and devices differed among studies, including conventional clips, clip-and-line methods, and other closure devices. Regarding DP, its incidence is considerably lower than that of DB and PECS; therefore, a larger number of cases may be required to clarify whether mucosal closure reduces its occurrence. Nevertheless, given the low incidence of DP, further large-scale prospective studies are warranted. Publication bias could not be completely excluded because of the limited number of studies for some outcomes.

 

2) Grouping of methods (clip suturing etc should be made in the text so that readers can more easily understand the information given), as was done in Table 2.

 

Answer: We really appreciate your valuable comments. According to the reviewer’s comments, we classified reports about closure techniques into 5 types as follows: 1. Clip+TIPS, 2. Clip+Line, 3. Clip+Device, 4. Special Clip, 5. Suturing device. We explained these types in the main text described below and adopted them into 22 reports in Table 2. Regarding Table 2, we divided it into two parts according to published year such as 2021-2020 (Table 2) and 2021-2026 (Table 3)

 

We classified reports about closure techniques into 5 types as follows: 1. Clip+TIPS, 2. Clip+Line, 3. Clip+Device, 4. Special Clip, 5. Suturing device. Clip+TIPS is the type which does not need any special devices but needs some tips such as incision, underwater situation, and muscle layer clipping. A representative report is modified double layer method (Origami method) [34]. Clip+Line is a clipping method with line and a representative report is the reopenable clip and line method (ROLM) [37]. Clip+Device is a clipping method with marketed accessory for other use and a representative report is clipping with SureClip Traction Band (Microtech, Nanjing, China) [41]. A special clip and suturing device are marketed ones for closure. Representatives are MANTIS Closure Device and Overstitch, respectively [21,38].

 

3) Discussion part could be enriched by incorporating some strategies/conclusions/decision making by the authors according to the evidence given in the aforementioned paragraphs.

 

Answer: Thank you for your important comments. We made a new paragraph 8 for discussing indication of closure, how to choose a type of closure and so on.

 

8. Indications for endoscopic closure based on reported evidence and clinical factors

Previous reviews and our simplified meta-analysis demonstrated that complete closure significantly reduces DB after colorectal ESD, providing strong evidence for its use [4–8]. However, the overall incidence of DB remains relatively low, occurring in 4.3% and 6.6% of patients in the closure and non-closure groups, respectively (Table 5). In contrast, certain high-risk factors substantially increase the risk of DB, including lesion size >50 mm (OR, 3.63), ASA class III–IV (OR, 2.26), and antithrombotic therapy (OR, 1.72) [60]. Our previous study also demonstrated the efficacy of complete closure in patients receiving anticoagulants, with DB rates of 5.2% and 10.8% in the closure and non-closure groups, respectively [54]. Therefore, complete closure should be strongly considered in high-risk cases, whereas its indication in low-risk cases may be deter-mined individually.

The choice of closure method should be based on defect size, cost, procedure time, operator expertise, and local availability. Most closure techniques have been evaluated for defects <50 mm (Table 2), where procedural efficiency and cost-effectiveness are important considerations. Our previous study reported a median closure time of 6.9 minutes for defects measuring 20–57 mm, suggesting that this method may be a practical option for defects <50 mm [43]. In contrast, only a limited number of techniques are suitable for defects >50 mm, including suturing devices, the Origami method, and ROLM.

Durable closure may be particularly important in patients receiving antithrombotic agents because DB tends to occur later in these patients [43]. Recent closure techniques using dedicated devices have achieved prolonged maintenance of defect closure. For example, MCD and the SureClip Traction Band maintained complete closure in all cases two days after ESD [38,41], while ROLM and OTSC have also demonstrated durable closure [65,66].

Although complete closure did not significantly reduce DP, this may be attributable to the low incidence of DP and insufficient statistical power. In clinical practice, we preferentially perform complete closure in lesions at high risk of DP, particularly large right-sided lesions and those requiring extensive coagulation near the muscle layer. However, both DP and PECS are partly caused by thermal injury to the muscle layer during dissection and hemostasis. Consistent with this mechanism, many studies have not demonstrated a significant reduction in PECS with complete closure (Table 3). Therefore, strategies aimed at minimizing thermal injury, such as lower electrosurgical settings and underwater techniques, may be more effective for preventing DP and PECS [67-69]. Furter studies are expected for proving it.

 

New reference

65. Kitahara G, Wada T, Murotani K, et al. Evaluation of Delayed Bleeding Prevention and Sustained Closure Using the Reopenable Clip-Over-the-Line Method for Gastric Endoscopic Submucosal Dissection. DEN Open. 2026; 6: e70302.

 

66. Armellini E, Crinò SF, Orsello M, et al. Novel endoscopic over-the-scope clip system. World J Gastroenterol. 2015; 21: 13587-13592.

 

67. Inoue K, Yoshida N, Kobayashi R, et al. Efficacy of gel immersion endoscopic submucosal dissection for

colorectal lesions. Endosc Int Open. Accepted on April 28 2026, In press.

 

4) How was table 4 constructed?was meta-analysis performed? Please provide more data and explanations

 

Answer: Thank you for your kind comment. It was due to the simplified meta-analysis which we explanted for your previous comment.

 

5) Line 302, authors state data taken together? Does this refer to meta-analysis?

 

Answer: We appreciate your comments. We amended this part and explained our simplified meta-analysis thoroughly as we showed in your previous comment.

 

6) Line 317, authors fail to explain how they reach those conclusions

 

Answer: Thank you for your comment. We showed a pooled analysis about the efficacy of closure for DP and it did not show a significant difference. However, as we mentioned the rate of DP is substantially lower than DB and PECS, referring our data and previous reports. So, we considered that our analysis might not have enough power for the proving the efficacy of closure for DP. We wrote this hypothesis in the limitation section which is made at the end of paragraph 6 newly described below.

 

Limitation section

Regarding DP, its incidence is considerably lower than that of DB and PECS. Therefore, the available evidence, including previous reviews and our simplified meta-analysis, may be insufficient to definitively determine the efficacy of complete closure for preventing DP. Further studies with larger sample sizes are needed to clarify this issue.

Reviewer 2 Report

Comments and Suggestions for Authors

This is a well‑structured and comprehensive narrative review on endoscopic closure after colorectal ESD, covering devices, techniques, clinical outcomes, and adverse‑event prevention. The manuscript is generally well-written, the tables and figures are informative, and the synthesis of recent studies up to 2026 is valuable for clinicians and endoscopists. The review succeeds in presenting a balanced overview of closure strategies and their relative strengths, and it will likely serve as a useful reference for practitioners.

Below are my comments to improve the manuscript.

1) The narrative could be improved by more explicitly comparing the relative evidence quality across methods (e.g., RCTs vs. retrospective cohorts)

2) The authors could better clarify which techniques are most supported in high‑risk populations, such as patients on antithrombotic therapy.

3)The section on PECS and delayed perforation would benefit from a more critical interpretation of why closure does not consistently reduce these events, including the role of thermal injury, lesion size, and operator‑dependent factors.

4) The availability of a different OTSC clip should be included, and a brief sentence added in the text. This clip has a different shape and design compared to the OVESCO, and the release system is also different. For the specific features of this device, you can refer to PMID 26730172.

5) A brief concluding paragraph highlighting practical recommendations for selective closure based on lesion and patient characteristics would enhance the clinical applicability of the review.

Author Response

 

Reviewer 2

This is a well‑structured and comprehensive narrative review on endoscopic closure after colorectal ESD, covering devices, techniques, clinical outcomes, and adverse‑event prevention. The manuscript is generally well-written, the tables and figures are informative, and the synthesis of recent studies up to 2026 is valuable for clinicians and endoscopists. The review succeeds in presenting a balanced overview of closure strategies and their relative strengths, and it will likely serve as a useful reference for practitioners.

 

Below are my comments to improve the manuscript.

 

1) The narrative could be improved by more explicitly comparing the relative evidence quality across methods (e.g., RCTs vs. retrospective cohorts)

 

Answer: We appreciate your comment. According to another reviewer’s comment, we added the simplified meta-analysis of the 22 reports in Table 3. In the meta-analysis, we did the analysis of the quality for RCTs and retrospective studies using the Newcastle-Ottawa scale and made Supplemental Table 1 [new reference].

 

Method

We searched PubMed, Embase, and the Cochrane Library for eligible studies up to April 2026 for a simplified meta-analysis about the efficacy of endoscopic closure. The key terms were (“closure” “suture” OR “Clip”) AND (“endoscopic submucosal dissection” OR “ESD”) AND (“Colorectum” “Colon” OR “Colorectal”) for examining the efficacy of various closure methods in this review. In this simplified meta-analysis, we evaluated the efficacy of endoscopic closure after colorectal ESD for preventing DB, DP, and PECS. The numbers of adverse events and total cases in the closure and non-closure groups were extracted from each eligible study. The definition of complete resection and each adverse events were followed by each report’s one. Inclusion criteria were (i) patients undergoing colorectal ESD; (ii) prophylactic endoscopic closure after ESD; (iii) existence of non-closure group; (iv) incidence of either DB, DP, and PECS; (v) randomized controlled trials (RCTs) or observational studies (prospective or retrospective, case-control, or cohort studies); (vi) publication in English. Two investigators (N.Y. and Y.T.) independently extracted the data and assessed study quality. Any discrepancies were re-solved by discussion.

 

Statistical analysis

Odds ratios (ORs) with 95% confidence intervals (CIs) were calculated for each outcome. When a study included a zero-event cell, a continuity correction of 0.5 was applied. Studies with no events in both groups were excluded from pooled effect estimation for the corresponding outcome. Pooled analyses were conducted using a random-effects model according to the DerSimonian–Laird method because clinical and methodological heterogeneity among studies was anticipated. Statistical heterogeneity was assessed using Cochran’s Q test and the I² statistic. An I² value >50% was considered indicative of substantial heterogeneity. Forest plots were generated for DB, DP, and PECS, including study weights, pooled estimates, and 95% CIs. Definitions of adverse events varied slightly among included studies, and the original definitions used in each study were accepted for analysis. All statistical analyses were performed using Python-based statistical scripts. Statistical analyses and forest plots were generated using Review Manager (RevMan) version 5.4 and Python version 3.10. Statistical heterogeneity was assessed using the I² statistic and Cochran’s Q test.

 

Results

A total of 520 studies were identified through database searching. After screening and eligibility assessment, 22 studies were included in the simplified meta-analysis (Table 4) [41,44-56]. This meta-analysis was performed for DB, DP, and PECS comparing closure and non-closure groups after colorectal ESD. The methodologic quality of the 4 RCTs and 18 retrospective studies were 7-8 and 5-8 according to the Newcastle-Ottawa scale (Supple-mental Table 1) [57].

Regarding the simplified meta-analysis, for DB, pooled analysis demonstrated that endoscopic complete closure was significantly associated with a reduced risk of DB (OR, 0.77; 95% CI, 0.60–0.97; p=0.030; I²=53.6%) (Figure 6). For DP, there was no significant difference between the closure and non-closure groups (OR, 0.60; 95% CI, 0.23–1.55; p=0.290; I²=0.0%) (Figure 7). For PECS, endoscopic closure was not significantly associated with risk reduction (OR, 0.94; 95% CI, 0.65–1.38; p=0.765; I²=56.6%) (Figure 8).

 

Figure 6

Forest plot of the included studies evaluating the efficacy of complete closure for DB after colorectal ESD

 

Figure 7

Forest plot of the included studies evaluating the efficacy of complete closure for DP after colorectal ESD

 

Figure 8

Forest plot of the included studies evaluating the efficacy of complete closure for PECS after colorectal ESD

 

Limitations

Several limitations should be considered in this meta-analysis. The definitions of DB, DP, and PECS were not completely uniform among the included studies, which may have contributed to heterogeneity. Most included studies were retrospective observational studies, and only a limited number of randomized controlled trials were available. Closure techniques and devices differed among studies, including conventional clips, clip-and-line methods, and other closure devices. Regarding DP, its incidence is considerably lower than that of DB and PECS; therefore, a larger number of cases may be required to clarify whether mucosal closure reduces its occurrence. Nevertheless, given the low incidence of DP, further large-scale prospective studies are warranted. Publication bias could not be completely excluded because of the limited number of studies for some outcomes.

 

Reference for the Newcastle-Ottawa scale

57. Stang A. Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur. J. Epidemiol. 2010; 25: 603–605.

 

2) The authors could better clarify which techniques are most supported in high‑risk populations, such as patients on antithrombotic therapy.

 

Answer: We appreciate your valuable comment. We made a new paragraph 8 for Indications for endoscopic closure based on reported evidence and clinical factors and explained expected techniques for high-risk populations.

 

Paragraph

8. Indications for endoscopic closure based on reported evidence and clinical factors

The indications for endoscopic closure after colorectal ESD remain highly controversial, as previous reviews and the simplified meta-analysis in the present study demonstrated a significant benefit only for delayed bleeding (DB) in the overall colorectal ESD population [4–8]. Careful consideration should therefore be given to patient- and lesion-related risk factors when determining whether closure should be performed, including antithrombotic therapy, lesion size, and colorectal location. Closure should be considered particularly for large tumors, right-sided colonic lesions that are more prone to delayed perforation, rectal lesions with a higher risk of delayed bleeding, and patients receiving anticoagulant therapy. Although various closure techniques have been developed, many conventional methods are technically feasible only for defects of ap-proximately 50 mm or smaller. For larger defects exceeding 50 mm, suturing devices or clip-and-line techniques such as ROLM may be more effective [37]. Another promising and convenient method is the Origami method [34]. In this technique, a first clip is deployed to muscle layer of the defect and the muscle layer was folded. The second clip grasped the anal side of the muscle layer, and another muscle was folded. Then these folded muscles were connected using a clip, and the mucosal defect was shrunk. These clippings are performed repeatedly and complete closure is achieved finally. Procedural speed for closure is also an important factor, and the recent development of specialized clips has contributed to shortening closure time. In particular, for defects of ap-proximately 50 mm or less, the MCD technique appears promising because of its particularly short closure time [43]. For high-risk patients, such as patients on antithrombotic therapy, our previous study showed the efficacy of endoscopic closure only for right-sided colon [54]. It suggests regular closure was not enough for preventing DB. Durable closure using recent specific clip and technique are expected. Closures with MCD and SureClip Traction band showed long-lasting closure [38,41]. ROLM also showed durable closure [63.1]. In addition, cost should also be taken into account, and closure strategies should be selected based on cost–benefit considerations. From this perspective, it is important to appropriately identify suitable indications for closure and to select the most appropriate devices according to the clinical situation.

 

3)The section on PECS and delayed perforation would benefit from a more critical interpretation of why closure does not consistently reduce these events, including the role of thermal injury, lesion size, and operator‑dependent factors.

 

Answer: We appreciate your comment. We think that thermal injury is a key factor for PECS and delayed perforation. We added the comment for it in the Paragraph 8.  Several papers showed the relationship between thermal injury and PECS. We referred these papers.

 

8. Indications for endoscopic closure based on reported evidence and clinical factors

Although complete closure did not significantly reduce DP, this may be attributable to the low incidence of DP and insufficient statistical power. In clinical practice, we preferentially perform complete closure in lesions at high risk of DP, particularly large right-sided lesions and those requiring extensive coagulation near the muscle layer. However, both DP and PECS are partly caused by thermal injury to the muscle layer during dissection and hemostasis. Consistent with this mechanism, many studies have not demonstrated a significant reduction in PECS with complete closure (Table 3). Therefore, strategies aimed at minimizing thermal injury, such as lower electrosurgical settings and underwater techniques, may be more effective for preventing DP and PECS [67-69]. Furter studies are expected for proving it.

 

New reference

65. Kitahara G, Wada T, Murotani K, et al. Evaluation of Delayed Bleeding Prevention and Sustained Closure Using the Reopenable Clip-Over-the-Line Method for Gastric Endoscopic Submucosal Dissection. DEN Open. 2026; 6: e70302.

 

66. Armellini E, Crinò SF, Orsello M, et al. Novel endoscopic over-the-scope clip system. World J Gastroenterol. 2015; 21: 13587-13592.

 

67. Inoue K, Yoshida N, Kobayashi R, et al. Efficacy of gel immersion endoscopic submucosal dissection for

colorectal lesions. Endosc Int Open. Accepted on April 28 2026, In press.

 

68. Dell'Unto E, Rimondi A, Kalopitas G, et al. The saline-immersion/irrigation technique (SITE) for colonic endoscopic submucosal dissection (ESD): A comprehensive evaluation of outcomes, efficacy, and safety. Saudi J Gastroenterol. 2025 Dec 3. doi: 10.4103/sjg.sjg_422_25. Epub ahead of print. PMID: 41335013.

 

69. Koyama Y, Fukuzawa M, Aikawa H, et al. Underwater endoscopic submucosal dissection for colorectal tumors decreases the incidence of post-electrocoagulation syndrome. J Gastroenterol Hepatol. 2023; 38:1566-1575.

 

4) The availability of a different OTSC clip should be included, and a brief sentence added in the text. This clip has a different shape and design compared to the OVESCO, and the release system is also different. For the specific features of this device, you can refer to PMID 26730172.

 

Answer: Thank you for your comment. We added the OTSC clip (Padlock Clip) according to your comment in Table 1 and added new reference about new OTSC.

 

Reference 66. Armellini E, Crinò SF, Orsello M, et al. Novel endoscopic over-the-scope clip system. World J Gastroenterol. 2015; 21: 13587-13592.

 

5) A brief concluding paragraph highlighting practical recommendations for selective closure based on lesion and patient characteristics would enhance the clinical applicability of the review.

 

Answer: We appreciate your comment. We made a new paragraph 8 about it described in the reply to your previous comment.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

I would like to congratulate the authors on the revised version of their work after reviewer's suggestions. I have no further comments to make

Author Response

Thank you for your comment.