Abstract
Introduction:
What interventions effectively prevent postoperative stenosis following endoscopic resection (ER) of superficial esophageal cancer? This study aimed to identify effective interventions or combinations through a systematic review and network meta-analysis.
Methods:
Six databases were systematically searched for eligible studies up to 30 April 2023, on interventions to prevent esophageal stenosis post-ER. Odds ratios (ORs) evaluated stenosis rate (primary outcome) and complications (secondary outcome), while mean differences (MD) evaluated endoscopic balloon dilatation (EBD) sessions post-stenosis.
Results:
Twenty-three studies involving 1271 patients and 11 different interventions were included. Eight interventions were effective in preventing post-ER stenosis: oral hydrocortisone sodium succinate and aluminum phosphate gel (OHA) (OR: 0.02, 95% credible interval [CrI]: 0.00–0.11), polyglycolic acid (PGA) + ST (OR: 0.02, 95% CrI: 0.00–0.23), oral tranilast (OT) + preemptive endoscopic balloon dilatation (PEBD) (OR: 0.08, 95% CrI: 0.01–0.77), botulinum toxin (BT) (OR: 0.10, 95% CrI: 0.03–0.32), ST (OR: 0.08, 95% CrI: 0.01–0.67), oral steroid (OS) (OR: 0.11, 95% CrI: 0.05–0.28), endoscopic triamcinolone injection (ETI) + OS (OR: 0.17, 95% CrI: 0.07–0.42), and ETI (OR: 0.18, 95% CrI: 0.11–0.30). Five interventions significantly reduced EBD sessions: PGA + ST (MD: −5.78, 95% CrI: −11.04 to −1.21), ETI + OS (MD: −3.27, 95% CrI: −5.37 to −0.72), OS (MD: −6.18, 95% CrI: −9.43 to −3.38), ETI (MD: −3.81, 95% CrI: −5.74 to −1.99), and BT (MD: −2.16, 95% CrI: −4.12 to −0.40). None of the interventions significantly increased complications.
Conclusions:
This study confirmed the efficacy of OS, ETI, and ETI + OS and verified five other interventions (OHA, PGA + ST, OT + PEBD, BT, and ST) in preventing stenosis. Notably, PGA + ST and BT also reduced the number of EBD sessions.
Keywords: endoscopic resection, esophageal cancer, esophageal stenosis, network meta-analysis, systematic review
Introduction
Esophageal cancer ranks the eighth in global cancer diagnoses and sixth in cancer-related deaths. Early-stage cases frequently lack typical symptoms and signs, leading to advanced-stage diagnoses for most patients[1]. Consequently, the 5-year survival rate for esophageal cancer remains low, typically between 10% and 30% in most countries[2]. Endoscopic resection (ER), including endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD), is widely acknowledged as a minimally invasive treatment for superficial esophageal cancer, boasting a 5-year survival rate exceeding 90%[3].
Despite its minimally invasive nature, ER can lead to esophageal stenosis in many patients post-procedure. Studies indicate that 11%–20% of patients develop esophageal stenosis following ESD[4]. The incidence of esophageal stenosis can range from 60% to 100% in patients with a postoperative mucosal defect circumference (MDC) greater than 3/4[5]. Although guidelines recommend proactive measures to prevent esophageal stenosis following ER for large esophageal lesions, standardized methods are still lacking[5,6]. The rapid advancement of medical technology continues to introduce new prevention strategies.
Previous pairwise meta-analyses have shown that steroids, polyglycolic acid (PGA), and stents are effective in preventing esophageal stenosis compared to placebos (PLA).[7–10] However, these meta-analyses did not include pairwise comparisons between interventions, limiting their ability to make definitive clinical decisions. In contrast, Bayesian network meta-analyses (NMA) can evaluate multiple interventions through both direct and indirect evidence. Even when direct evidence appears conclusive, NMA can provide a more precise estimation of the effects of different interventions[11].
Therefore, we conducted a systematic review to comprehensively gather clinical evidence and performed an NMA to compare and rank various interventions. This NMA aimed to identify the most effective strategy for preventing esophageal stenosis and maximizing patient benefit.
Methods
This systematic review and NMA rigorously followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses and the Assessing the Methodological Quality of Systematic Reviews guidelines[12,13]. The protocol was registered in PROSPERO. Complete details are accessible online, with any protocol amendments noted in the Supplementary Digital Content, Appendix (Appendix 18, http://links.lww.com/JS9/D771).
Search strategy
We conducted a thorough and methodical search across six databases (Cochrane, Embase, PubMed, Web of Science, Scopus, and CINAHL) to identify published randomized clinical trials (RCTs) and observational studies (cohort and case-control studies) from their inception to 30 April 2023. In addition, we manually reviewed reference lists of relevant studies and international trial registers for additional related research. Search terms included “esophageal,” “stricture,” “carcinoma,” “endoscopic submucosal dissection,” “endoscopic mucosal resection,” among others. Only publications in English were considered. The detailed search strategy is provided in Supplementary Digital Content, Appendix 1 (http://links.lww.com/JS9/D771). Y.D. and H.X. independently conducted and cross-checked the search, and summarized the findings.
Selection criteria
The complete PICOS criteria for the NMA were as follows: (1) Patients: Patients were diagnosed with superficial esophageal cancer and underwent ESD or EMR involving at least 50% of the mucosal circumference. (2) Intervention: Interventions included botulinum toxin (BT) injection, endoscopic triamcinolone injection (ETI), oral steroid (OS), oral hydrocortisone sodium succinate and aluminum phosphate gel (OHA), oral tranilast (OT), preemptive endoscopic balloon dilatation (PEBD), PGA, stent (ST) placement, and combinations of two interventions (ETI + OS, ETI + PGA, OT + PEBD, PGA + ST). To ensure adequate direct comparisons for the meta-analysis, interventions were clustered based on treatment modality despite variations in dosage (OS and OHA), injection method (ETI), and treatment duration (OS). (3) Control: PLA/no treatment (PLA). (4) Outcomes: The primary outcome was the esophageal stenosis rate after ER. Secondary outcomes included the number of endoscopic balloon dilatation (EBD) sessions required to treat stenosis and intervention-related complications (detailed definitions in Supplementary Digital Content, Appendices 3 and 4, http://links.lww.com/JS9/D771). (5) Study designs: RCTs and observational studies.
The inclusion criteria were as follows: 1) Patients over 18 years with superficial esophageal cancer who underwent ER (ESD or EMR) involving at least 50% of the MDC. 2) Absence of lymph node metastasis confirmed by endoscopic ultrasonography or computed tomography. 3) Studies must have reported at least the primary outcome. 4) Due to the inclusion of case-control studies, there was no restriction on the follow-up duration.
Exclusion criteria for the studies included: 1) guidelines, expert positions or consensus, brief reports, case reports, letters, comments, protocol studies, reviews, or meta-analyses; 2) single-arm trials, pharmacokinetic studies, animal or cell studies; 3) studies reporting different outcomes of interest; 4) studies with incomplete or duplicate data. 5) Studies lacking full-text availability. 6) Studies where the case and control groups received the same intervention (e.g., OS for 3 vs. 6 weeks). 7) Studies with an overall sample size of fewer than 10 patients or failure to complete statistical analysis.
Two reviewers (Y.D. and Y.X.) independently assessed titles and abstracts. For studies meeting the inclusion criteria, we thoroughly reviewed of the full text. Any discrepancies between the two reviewers were resolved through consensus or escalated to other reviewers for resolution (W.Y., G.Z., and Z.L.).
Data extraction
Four independent reviewers (Y.D., Z.Z., S.X., and H.X.) conducted the data extraction. This process involved collecting information on the first author, year of publication, study design, sex, median age, intervention, MDC, endoscopic treatment, follow-up duration, and the total number of treated patients from the included studies. Patients with an MDC of less than 3/4 were excluded from the primary outcome analysis. However, for secondary outcomes, these patients were not excluded due to insufficient data in the original studies.
Methodological quality
The Cochrane Risk-of-Bias Tool for Randomized Trials (RoB 2) was employed to evaluate the risk of bias using Review Manager software (Version 5.3). This tool assesses six components: selection bias, performance bias, detection bias, attrition bias, reporting bias, and other sources of bias[14]. Studies with a high risk of bias in one or more components were categorized as high risk overall. The Newcastle–Ottawa Scale was utilized to assess the quality of observational studies. Studies scoring <6 stars on the scale were considered to have a high risk of bias (http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp). In addition, the quality of evidence contributing to each network estimate was evaluated using the grading of recommendations assessment, development, and evaluation (GRADE) framework, which characterizes the quality of evidence based on study limitations, imprecision, inconsistency, indirectness, and publication bias for each outcome[15].
Publication bias
When the number of included studies exceeded 10, publication bias was evaluated both qualitatively and quantitatively[16]. Qualitative assessment involved examining a funnel plot, where the symmetry of the plot indicated the presence or absence of publication bias. The quantitative assessment was performed using Egger’s and Begg’s tests[17].
Statistical analysis
First, we conducted an NMA using the statistical package “multinma” in R (version 4.2.3)[18]. Both fixed-effect and random-effect models were conducted, and we compared them based on leverage effect, deviance information criterion, and residual deviance to determine the model with a better fit. The Markov Chain Monte Carlo method was used to obtain pooled results. In random state selection, three Markov chains ran simultaneously. All comparisons assumed a common heterogeneity parameter[19]. For binary outcomes, pooled results were presented as odds ratio (OR). Continuous outcomes used mean differences (MD) instead of standardized MD due to consistent units and measurement of EBD data. All results were reported with a 95% confidence interval (CI) for direct comparison or credible interval (CrI) for NMA[20]. Analyses for RCTs were conducted on an intention-to-treat (ITT) basis rather than per-protocol analysis to mitigate the risk of type I error, as ITT analysis is more conservative[21]. The Cochran’s Q test and I2 statistic, calculated using Stata 17.0, were used to evaluate heterogeneity, with I2 > 50% and P < 0.1 indicating substantial heterogeneity[22].
Second, we evaluated the inconsistency between direct and indirect evidence sources of evidence globally (by comparing the accuracy and fit of the consistency and inconsistency models) and locally (by calculating the differences between direct and indirect estimates in all closed loops within the network)[23]. We used the node-splitting method to quantify model inconsistency[24].
Third, we estimated the probability of treatment rankings and generated surface under the cumulative ranking curves (SUCRA) to illustrate the cumulative ranking probability maps for different interventions[25]. We also conducted cluster analyses to find optimal treatment clusters based on two outcomes[23].
Fourth, to assess the influence of study characteristics on the results, we conducted subgroup NMA for the primary outcome based on the study type, follow-up duration, and extent of mucosal circumference. Moreover, we conducted sensitivity NMA for the primary outcomes by excluding studies with extreme results, case-control studies, or those involving EMR to validate the robustness of the results.
Results
Study inclusions
A total of 7350 articles were initially identified through the search process. After removing 4079 duplicates, we carefully reviewed the titles and abstracts of the remaining 3271 studies. We then assessed the full text of 194 studies for eligibility. Out of these, 167 studies were excluded for the following reasons: 33 were brief reports, 17 involved patients not undergoing ER, 41 were single-arm trials, 4 were single intervention control trials, 30 were animal trials, 4 were cell trials, 30 were protocol studies, 2 lacked a primary outcome, 5 were meta-analyses, and 1 involved duplicate data. Finally, 6 RCTs, 16 cohort studies, and 1 case-control study were included in the NMA (Fig. 1).
Figure 1.
Flowchart of study selection.
Characteristics and quality assessment of the studies
A total of 1271 patients were included, with 1097 having an MDC >3/4. The patients’ ages ranged from 41 to 90 years, with a median age of 66.7 years. The proportion of male patients varied from 59.1% to 97.4%, with a median of 79.8%. Thirteen studies (57%) included patients from Japan, nine (39%) from China, and one (4%) from Korea. Only two studies included patients undergoing EMR, while the others included only those undergoing ESD. Follow-up duration ranged from 1.5 to 84 months, with a median of 9.5 months (Table 1). Regarding quality, two RCTs (33%) were rated as high risk of bias (Supplementary Digital Content, Appendix 5, http://links.lww.com/JS9/D771), and all observational studies scored above six points (Table 2).
Table 1.
Feature of included studies.
| Publication | Country | Design | Age (range) | Proportion of male (%) | Mucosal defect circumference | Intervention | Control | Case number (protocol:control) | Stricture number (protocol:control) | Outcome | Treatment | Follow-up time (month) | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Protocol 1 | Protocol 2 | Incidence of stricture (protocol:control) | EBD number (protocol:control) | Complication number (protocol:control) | |||||||||||
| Ezoe 2011 | Japan | Cohort study(retrospective) | 62 (42–80) | 95.1% | >3/4 | PEBD | PLA | 29/12 | 17/11 | 58.6%/91.7% | 2.0 (2–20) 4.5 (2–35) | 0/1 | ESD/EMR | 84 (median) | |
| Hashimoto 2011 | Japan | Cohort study (retrospective) | 70.9 (49–89) | 87.8% | >3/4 | ETI | PLA | 21/20 | 4/15 | 19.0%/75.0% | 1.7 (0–15) 6.6 (0–20) | 0/0 | ESD | 24 | |
| Yamaguchi 2011 | Japan | Cohort study (retrospective) | 68.4 (51–82) | 85.4% | >3/4 | OS | PEBD | 19/22 | 1/7 | 5.3%/31.8% | 1.7 (0–7) 6.8 (0–32) | 0/1 | ESD | 12 (median) | |
| Uno 2012 | Japan | RCT | 68.4 (NA) | 93.5% | >3/4 | OT + PEBD | PEBD | 15/16 | 5/11 | 33.3%/68.8% | 0 (0–1.75) 4.0 (0–6.5) | 0/1 | ESD | 26 (median) | |
| Hanaoka 2012 | Japan | Cohort study (prospective) | 65.5 (NA) | 89.8% | >3/4 | ETI | PLA | 30/29 | 3/19 | 10.0%/65.5% | 0 (0–2) 2 (2–15) | 2/0 | ESD | 2 | |
| Kataoka 2014 | Japan | Cohort study (retrospective) | 69.3 (57–85) | 78.8% | >3/4 | OS | PLA | 17/16 | 3/11 | 17.6%/68.7% | 4.6 (2–10) 8.1 (1–18) | 0/0 | ESD | 12 (median) | |
| Wen 2014 | China | RCT | 61.4 (NA) | 59.1% | >3/4 | ST | PLA | 11/11 | 2/8 | 18.2%/78.7% | 0.9 ± 0.9 5.7 ± 5.3 | 2/0 | ESD | 3 | |
| Takahashi 2015 | Japan | RCT | 70.5 (48–89) | 78.8% | >3/4 | ETI | PLA | 16/16 | 10/14 | 62.5%/87.5% | 6.1 ± 6.2 12.5 ± 10.1 | 0/0 | ESD | 16 | |
| Nagami 2015 | Japan | Case–control | 69.5 (NA) | 82.1% | >2/3 | ETI | PLA | 28/28 | 3/10 | 10.7%/35.7% | NA | 0/0 | ESD | NA | |
| >3/4 | ETI | PLA | 17/19 | 2/10 | 11.8%/52.6% | NA | |||||||||
| Kadota 2016 | Japan | Cohort study (retrospective) | 70 (NA) | 86.1% | >3/4 | ETI + OS | ETI | PLA | 29/53/33 | 12/23/22 | 41.4%/43.4%/66.7% | 5.5 (4–8.3) 6.0 (4.5–10.5) 12.5 (7.5–16) | 1/2/0 | ESD/EMR | 12 |
| Wen 2016 | China | RCT | 60.9 (NA) | 68.7% | >1/2 | BT | PLA | 33/34 | 2/11 | 6.1%/32.4% | 1.5 (0–2) 2.8 (0–5) | 0/0 | ESD | 3 | |
| >3/4 | BT | PLA | 8/11 | 2/7 | 25.0%/63.6% | NA | 0/0 | ESD | 3 | ||||||
| Tsujii 2017 | Japan | Cohort study (retrospective) | 71 (52–83) | 97.4% | >3/4 | ETI | PLA | 28/10 | 12/9 | 42.9%/90.0% | 8 (1–186) 5 (2–12) | 5/1 | ESD | 2 | |
| Zhou 2017 | China | Cohort study (retrospective) | 66.4 (NA) | 65.2% | >3/4 | OS | PLA | 13/10 | 3/8 | 23.1%/80.0% | 0.69 (0–3) 13.5 (0–28) | 0/0 | ESD | 12 | |
| Iizuka 2017 | Japan | Cohort study (retrospective) | 67.9 (NA) | 78.6% | >2/1 | PGA | ETI | 33/29 | 3/3 | 9.1%/10.3% | 0.057 ± 0.24 1.9 ± 5.1 | 1/0 | ESD | 1.5 | |
| >3/4 | PGA | ETI | 16/12 | 3/1 | 18.6%/8.3% | NA | |||||||||
| Chai 2018 | China | RCT | 61.3 (NA) | 60.6% | >3/4 | PGA + ST | ST | 34/32 | 7/15 | 20.5%/46.7% | 4 (2–5) 6 (1–14) | NA | ESD | 2 | |
| Pih 2019 | Korea | Cohort study (retrospective) | 66 (NA) | 96.2% | >3/4 | OS | ETI | PLA | 25/6/22 | 5/2/11 | 20.0%/33.3%/50.0% | NA | NA | ESD | 20 (median) |
| Chu 2019 | China | Cohort study (retrospective) | 64.3 (46–80) | 65.7% | >2/3 | ETI + OS | PLA | 34/36 | 5/19 | 14.7%/51.5% | 0.2 ± 0.6 3.3 ± 5.4 | 1/1 | ESD | 28 (median) | |
| >3/4 | ETI + OS | PLA | 11/18 | 2/15 | 18.2%/83.3% | 0.4 ± 1.0 6.1 ± 6.5 | |||||||||
| Yang 2019 | China | Cohort study (prospective) | 62.8 (NA) | 66.7% | >3/4 | PGA + ST | ST | 38/37 | 5/13 | 13.2%/35.1% | 4.4 ± 0.9 5.8 ± 2.9 | NA | ESD | 2 | |
| Nie 2019 | China | Cohort study (retrospective) | 66.6 (NA) | 88.9% | >3/4 | OHA | ETI + OS | 14/13 | 1/7 | 7.1%/53.8% | 0 0.5 (0–1) | 9/9 | ESD | 2 | |
| Hashimoto 2019 | Japan | Cohort study (retrospective) | 71.5 (49–90) | 86.2% | >3/4 | ETI | PLA | 35/23 | 16/17 | 45.7%/73.9% | 0 (0–7) 4 (0–20) | 2/1 | ESD | 12 | |
| Sakaguchi 2020 | Japan | Cohort study (retrospective) | 69.3 (NA) | 85.9% | >3/4 | ETI + PGA | PGA | PLA | 49/30/32 | 18/12/17 | 36.7%/40.0%/53.1% | 4.8 ± 7.8 3.5 ± 6.7 5.2 ± 7.9 | 1/0/0 | ESD | 44 |
| Zhou 2021 | China | Cohort study (prospective) | 65.2 (NA) | 67.9% | >2/3 | BT | ETI | PLA | 26/16/36 | 7/7/30 | 26.9%/43.8%/83.3% | 1.19 (0–12) 1.31 (0–9) 3.14 (0–16) | 1/1/4 | ESD | 12 |
| >3/4 | BT | ETI | PLA | 17/12/27 | 7/7/25 | 41.2%/58.3%/83.3% | NA | ||||||||
| Zhang 2022 | China | RCT | 65 (41–85) | 77.8% | >3/4 | OHA | ETI + OS | 32/31 | 3/11 | 9.4%/35.5% | 2 (2–2) 2 (1–4) | 8/6 | ESD | 15 (median) | |
BT, botulinum toxin; EBD, endoscopic balloon dilatation; ETI, endoscopic triamcinolone injection; EMR, endoscopic mucosal resection; ESD, endoscopic submucosal dissection; NA, not available; OS, oral steroid; OT, oral tranilast; OHA, oral hydrocortisone sodium succinate and aluminum phosphate gel; PEBD, preemptive endoscopic balloon dilatation; PGA, polyglycolic acid; PLA, placebo/no treatment; RCT, randomized clinical trial; ST, stent.
Table 2.
Newcastle–Ottawa Scale scores of included studies.
| Selection | Comparability | Outcome | |||||||
|---|---|---|---|---|---|---|---|---|---|
| Study | Representativeness of the exposed cohort | Selection of the nonexposed cohort | Ascertainment of exposure | Demonstration that outcome of interest was not present at start of study | Comparability of cohorts on the basis of the design or analysis | Assessment of outcome | Was follow-up long enough for outcomes to occur | Adequacy of follow-up of cohorts | Total scores |
| Ezoe 2011 | ★ | ★ | ★ | ★ | ★ | ★ | ★ | 7 | |
| Hashimoto 2011 | ★ | ★ | ★ | ★ | ★★ | ★ | ★ | 8 | |
| Yamaguchi 2011 | ★ | ★ | ★ | ★ | ★ | ★ | ★ | 7 | |
| Hanaoka 2012 | ★ | ★ | ★ | ★ | ★★ | ★ | ★ | ★ | 9 |
| Kataoka 2014 | ★ | ★ | ★ | ★ | ★ | ★ | ★ | ★ | 8 |
| Nagami 2015 | ★ | ★ | ★ | ★★ | ★ | 6 | |||
| Kadota 2016 | ★ | ★ | ★ | ★ | ★ | ★ | ★ | 7 | |
| Tsujii 2017 | ★ | ★ | ★ | ★ | ★★ | ★ | ★ | ★ | 9 |
| Zhou 2017 | ★ | ★ | ★ | ★★ | ★ | ★ | ★ | 8 | |
| Iizuka 2017 | ★ | ★ | ★★ | ★ | ★ | 6 | |||
| Pih 2019 | ★ | ★ | ★ | ★ | ★★ | ★ | ★ | 8 | |
| Chu 2019 | ★ | ★ | ★★ | ★ | ★ | ★ | 7 | ||
| Yang 2019 | ★ | ★ | ★ | ★★ | ★ | 6 | |||
| Nie 2019 | ★ | ★ | ★ | ★ | ★★ | ★ | ★ | ★ | 9 |
| Hashimoto 2019 | ★ | ★ | ★ | ★ | ★★ | ★ | ★ | 8 | |
| Sakaguchi 2020 | ★ | ★ | ★ | ★ | ★ | ★ | 6 | ||
| Zhou 2021 | ★ | ★ | ★ | ★★ | ★ | ★ | ★ | 8 | |
Network structure and results
Figure 2 illustrates the network of eligible comparisons for stenosis rate. We included 11 interventions, with 8 of them directly compared to PLA. For graphical representations of the EBD sessions and complications, refer to Supplementary Digital Content, Appendix 7, http://links.lww.com/JS9/D771. Detailed results of the direct pairwise meta-analyses are presented in Supplementary Digital Content, Appendix 8, http://links.lww.com/JS9/D771. In the direct comparisons, ETI (OR: 0.19, 95% Cl: 0.11–0.31), OS (OR: 0.15, 95% Cl: 0.06–0.36), and BT (OR: 0.09, 95% Cl: 0.03–0.35) were effective in reducing the stenosis rate after ESD compared to PLA. In addition, OHA was more effective than ETI + OS (OR: 0.14, 95% CI: 0.04–0.47), and ST alone was less effective than ST + PGA (OR: 0.02, 95% CI: 0.00–0.11) in preventing stenosis.
Figure 2.
Network of eligible comparisons for stenosis rate. BT, botulinum toxin; ETI, endoscopic triamcinolone injection; OS, oral steroid; OT, oral tranilast; OHA, oral hydrocortisone sodium succinate and aluminum phosphate gel; PEBD, preemptive endoscopic balloon dilatation; PGA, polyglycolic acid; PLA, placebo/no treatment; ST, stent.
The results of the NMA for the primary outcome are presented as a league table in Figure 3. Interventions were ranked by efficacy according to SUCRAs. 1) Eight interventions were identified to prevent post-ER stenosis: OHA (OR: 0.02, 95% Crl: 0.00–0.11), PGA + ST (OR: 0.02, 95% Crl: 0.00–0.23), OT + PEBD (OR: 0.08, 95% Crl: 0.01–0.77), BT (OR: 0.10, 95% Crl: 0.03–0.32), ST (OR: 0.08, 95% Crl: 0.01–0.67), OS (OR: 0.11, 95% Crl: 0.05–0.28), ETI + OS (OR: 0.17, 95% Crl: 0.07–0.42), and ETI (OR: 0.18, 95% Crl: 0.11–0.30). 2) Five interventions outperformed PGA: OHA (OR: 0.04, 95% Crl: 0.01–0.26), PGA + ST (OR: 0.04, 95% Crl: 0.00–0.51), BT (OR: 0.18, 95% Crl: 0.04–0.83), OS (OR: 0.20, 95% Crl: 0.05–0.78), and ETI (OR: 0.32, 95% Crl: 0.11–0.96). 3) Four interventions were notably more effective than ETI + PGA: OHA (OR: 0.05, 95% Crl: 0.01–0.30), PGA + ST (OR: 0.05, 95% Crl: 0.00–0.57), BT (OR: 0.20, 95% Crl: 0.04–0.93), and OS (OR: 0.22, 95% Crl: 0.06–0.86). 4) OHA also exhibited superiority over ETI + OS (OR: 0.14, 95% Crl: 0.04–0.49), ETI (OR: 0.13, 95% Crl: 0.03–0.61), and PEBD (OR: 0.07, 95% Crl: 0.01–0.65). 5) ST + PGA showed significant superiority to ST alone (OR: 0.029, 95% Crl: 0.12–0.68).
Figure 3.
The pairwise network meta-analysis result for stenosis rate. Interventions are reported in order of SUCRA ranking. Comparisons should be read from left to right. The estimate is located at the intersection of the column-defining intervention and the row-defining intervention. BT, botulinum toxin; CrI, credible interval; ETI, endoscopic triamcinolone injection; OHA, oral hydrocortisone sodium succinate and aluminum phosphate gel; OR, odds ratio; OS, oral steroid; OT, oral tranilast; PEBD, preemptive endoscopic balloon dilatation; PGA, polyglycolic acid; PLA, placebo/no treatment; ST, stent; SUCRA, the surface under the cumulative ranking curve.
The NMA results for secondary outcomes are presented in Supplementary Digital Content, Appendix 8 (http://links.lww.com/JS9/D771). Compared to PLA, the following interventions significantly reduced EBD sessions: PGA + ST (MD: −5.78, 95% CrI: −11.04 to −1.21), ETI + OS (MD: −3.27, 95% CrI: −5.37 to −0.72), and OS (MD: −6.18, 95% CrI: −9.43 to −3.38), ETI (MD: −3.81, 95% CrI: −5.74 to −1.99), and BT (MD: −2.16, 95% CrI: −4.12 to −0.40). OS also significantly reduced the EBD sessions compared to BT (MD: −4.02, 95% CrI: −7.97 to −0.59). The top-ranked interventions for reducing EBD sessions, based on SUCRA scores, were: PGA + ST (SUCRA score, 81.5), ETI + OS (SUCRA score, 71.6), and OHA (SUCRA score, 71.5). Regarding complications, no significant differences were observed in comparisons between interventions and PLA, or in any pairwise comparisons.
Assessment of inconsistency and publication bias
The global, loops, and node-splitting tests did not reveal any inconsistencies in the stenosis rate and complications outcomes (all P-values >0.05), except in EBD sessions (global inconsistency P-value <0.001). However, the loop inconsistency tests did not detect a clear source of inconsistency in EBD sessions (all P-values >0.05). Further node-splitting analysis indicated that the inconsistency stemmed from the comparison between ETI + OS and PLA (P-value <0.001).
Subgroup, sensitivity, and cluster analyses
In the subgroup analysis based on MDC, both OHA and ETI + OS were effective in preventing stenosis in the MDC = 1 and 3/4 < MDC < 1 groups, while OS, BT, and ETI were only effective in the 3/4 < MDC < 1 group. In the subgroup analysis based on study type, OHA, ETI + OS, OS, BT, and ETI were effective in observational studies, but none were effective in RCTs.
To minimize the influence of endoscopic surgical methods, patients undergoing EMR were excluded. Sensitivity analysis showed that, except for OT + PEBD, the other interventions remained effective in preventing stenosis. OHA and PGA also reduced the number of EBD sessions. Additional sensitivity analyses, excluding case-control studies, confirmed that effective interventions for stenosis rate remained consistent. OHA became a significant intervention for reducing EBD sessions, while PGA + ST lost significance. Further sensitivity analyses, excluding studies with extreme values or with MDC <3/4, confirmed that the results were generally robust.
Cluster analyses grouped interventions based on the primary outcome (stenosis rate) and secondary outcomes (EBD sessions or complications). In Figure 4A, the 11 interventions are ranked according to SUCRA values for stenosis rate and EBD sessions, while in Figure 4B, they are ranked based on stenosis rate and complications. Considering both stenosis rate and EBD sessions, PGA + ST and OHA are potential optimal interventions. When considering both stenosis rate and complications, OHA and OT + PEBD emerge as potential optimal interventions.
Figure 4.
Clustered ranking plots based on cluster analysis of SUCRA values. (A) For stenosis rate and EBD sessions. (B) For stenosis rate and complication rate. Each color represents a group of interventions that belong to the same cluster. Interventions lying in the upper right corner are more optimal than the other prevention. BT, botulinum toxin; ETI, endoscopic triamcinolone injection; OS, oral steroid; OT, oral tranilast; OHA, oral hydrocortisone sodium succinate and aluminum phosphate gel; PEBD, preemptive endoscopic balloon dilatation; PGA, polyglycolic acid, PLA, placebo/no treatment; ST, stent.
Quality-of-evidence assessment
The GRADE framework assessed comparisons between each intervention and PLA, direct comparisons, and statistically significant comparisons in the NMA. It was also applied to assess overall rankings based on SUCRAs (Supplementary Digital Content, Appendix 16, http://links.lww.com/JS9/D771). The quality of evidence for each comparison initially started at a high level; however, it was downgraded due to factors such as risk of bias, inconsistency, indirectness, imprecision, and publication bias. According to GRADE, the quality of evidence for most comparisons regarding stenosis rate was rated very low, while the quality for most comparisons concerning EBD sessions was rated low. The overall quality of evidence was very low for ranking interventions in terms of stenosis rate and complications, and low for EBD sessions.
Discussion
Postoperative esophageal stenosis is a significant challenge for both patients and clinicians, particularly for patients with early-stage esophageal cancer who may not display overt symptoms. The reported incidence of stenosis varies widely due to the different studies considered in various guidelines. In this NMA, we recalculated the incidence of esophageal stenosis after ER to be 73.1% without preventive interventions and 33% with various interventions. When the MDC was 100%, the stenosis rate reached 100%, and even with interventions, it remained as high as 61%. However, the stenosis rate with interventions in our study should be interpreted cautiously, as our calculations did not include single-arm studies lacking a control group.
The pathogenesis of esophageal stenosis primarily involves structural damage to the esophageal epithelial barrier, chronic inflammation, and severe fibrosis[26]. Steroids are classic anti-inflammatory drugs, and OS and ETI are primary interventions to prevent esophageal stenosis. OHA, a combination of hydrocortisone sodium succinate and aluminum phosphate gel, acts as a barrier on artificial ulcers and creates a localized microenvironment enriched with hydrocortisone to suppress inflammation[27]. Positive results from small-sample studies, combined with its oral administration and cost-effectiveness, suggest that OHA holds promise as a steroid-based intervention. Tranilast inhibits the release of histamine and prostaglandins from mast cells, as well as fibroblast proliferation and collagen synthesis[28]. BT reduces muscle contraction and inhibits collagen fiber deposition and fibrous connective tissue formation[29]. PGA forms a protective layer over wounds and acts as a biophysical barrier, also providing a scaffold for cell adhesion and proliferation to reduce scar contracture[9]. In addition, physical dilation techniques (PEBD and ST) have also been used to prevent stenosis[10,30].
Utilizing an NMA with the GRADE methodology to critically assess and inform the evidence, we have highlighted key perspectives previously emphasized in guidelines or reviews. 1) OS and ETI, the two most commonly utilized interventions, showed no significant differences in preventing stenosis or reducing EBD sessions. The primary distinction between OS and ETI is the method of steroid administration (systemic vs. local). The combination of OS and ETI did not exceed the efficacy of either intervention alone. Given that ETI may weaken the esophageal wall, increasing its susceptibility to dilation pressures and potentially causing delayed perforation due to muscle layer damage[31], and considering the higher ranking of OS in our NMA, we recommend OS over ETI as the preferred intervention. 2) The combination of PGA and ST was more effective than either PGA or ST alone. ST supports the PGA sheet on the wound surface, preventing detachment. In addition, this combination reduced the risk of ST migration[32]. 3) The efficacy of PEBD in preventing stenosis appears limited. Although some guidelines recommend PEBD, the supporting evidence is solely based on a retrospective study with a small sample size (41 patients) conducted by Ezoe et al. This study included only a few patients who underwent ESD (5 cases) compared to 36 EMR cases, potentially introducing significant bias[30]. Considering the risks of bleeding, perforation before stenosis formation, and the additional psychological and economic burden on patients, we do not recommend PEBD for stenosis prevention. Notably, the Japanese ER guidelines (second edition, 2019) have also ceased recommending PEBD[5]. 4) The complication rates observed in this NMA were 0% in the OS group, 5.4% in the ETI group, and 19% in the ETI + OS group. We acknowledge that inconsistencies in defining and reporting steroid-related complications might have impacted the results of this NMA. Reported complications primarily included infections and post-EBD perforations rather than osteoporosis, diabetes, peptic ulcers, and psychiatric disorders noted by some scholars. The included studies did not report these specific complications, potentially because they are more commonly associated with prolonged, high-dose intravenous steroid use. This NMA suggests that while concerns regarding steroid-related complications are warranted, they may be overestimated. Short-term OS and local ETI appear to be relatively safe.
This NMA conducted multiple sensitivity and subgroup analyses to verify the effectiveness of each intervention. As none of the included studies reported EBD sessions based on MDC, we were unable to conduct further subgroup analysis for EBD sessions based on MDC. Therefore, we excluded studies with MDC <3/4 in the sensitivity analysis to minimize MDC-related heterogeneity. Considering the potential impact of study design, we excluded one case-control study in the sensitivity analysis and performed subgroup analyses of RCTs and cohort studies. However, some interventions were effective in observational studies but not in RCTs, likely due to the limited number and sample size of included RCTs, with most interventions reported in only one RCT. In addition, two studies involving patients with EMR were excluded to reduce heterogeneity from endoscopic techniques. Notably, most sensitivity analysis results were stable. The fluctuation observed in the OT + PEBD stemmed from the loss of important direct comparison after excluding studies involving EMR. After excluding studies with MDC <3/4, OHA became effective in reducing EBD sessions.
Based on our findings, we propose several recommendations for future studies: 1) Future studies should conduct direct comparisons between PGA + ST and OHA to identify the optimal intervention. Our NMA revealed that these two interventions had similar ORs (vs. PLA), SUCRA values, and cumulative rankings, with no significant differences in their pairwise comparison. Considering the high incidence of stenosis, we advise against using PLA controls in future studies; instead, head-to-head comparisons would be more ethical and appropriate. 2) Our study focused exclusively on patients with MDC >3/4 to reduce heterogeneity. However, this may limit the external validity of our results. European and Japanese guidelines recommend initiating prevention when MDC >1/2. Future studies should assess the cost-effectiveness of preventing stenosis in cases with MDC ranging from 1/2 to 3/4. 3) While Zhang et al. reported that most esophageal stenosis occurs within 4 weeks post-ESD[27], our NMA found a stenosis rate of 34.1% in studies with a median follow-up of less than 12 months, compared to 47.5% in studies with follow-ups exceeding 12 months. Therefore, future research should include follow-ups extending beyond 12 months and investigate optimal follow-up strategies after ESD. 4) We anticipate the emergence of large-sample RCTs comparing commonly used interventions with those previously reported in small-sample, single-arm studies, such as mitomycin C, self-help inflatable balloons, and biodegradable stents.[33–35] In addition, we look forward to validating autologous tissue transplantation and regenerative medicine through clinical trials[36,37].
This NMA represents the most comprehensive synthesis of data regarding current clinical interventions for preventing esophageal stenosis after ER. However, certain limitations related to both the NMA and individual studies warrant further discussion. First, due to the lack of direct head-to-head trials for certain comparisons, the quality of evidence within the GRADE framework ranges from low to very low. This limitation impacts the strength of the conclusions drawn from the NMA. Second, potential biases from selective reporting and small sample sizes in some studies limit the clinical interpretation of the findings. Despite extensive efforts to retrieve unpublished data and to contact study authors for supplementary data, some unpublished or published studies may still be missing, which could lead to an overestimation of the efficacy of certain interventions. Third, our NMA included a limited number of RCTs. Designing RCTs in endoscopic studies is challenging due to difficulties in blinding operators and randomizing patients, which introduces common biases. In addition, most patients in the control group may be unwilling to accept the poor quality of life associated with dysphagia caused by stenosis. Fourth, we could not provide detailed recommendations regarding optimal dosages, injection methods, and tapering regimens of steroids due to the limited number of eligible studies. Further research is needed to determine the optimal duration, dosage, and tapering regimen of OS, as well as the injection method, frequency, dose, and concentration of triamcinolone acetonide in ETI.
In conclusion, this NMA systematically reviewed and synthesized data from controlled clinical studies on 11 interventions for preventing postoperative stenosis after ER of superficial esophageal cancer. Eight interventions were found effective in preventing stenosis, and five significantly reduced the number of EBD sessions required post-stenosis. Notably, OHA and PGA + ST emerged as more promising than traditional steroid-based prevention methods. Future research should focus on large-scale, multicenter studies to comprehensively assess the effectiveness and safety of these emerging interventions. Advances in tissue engineering and regenerative medicine offer promising avenues for addressing post-ER stenosis.
Acknowledgements
We would like to thank Editage (www.editage.cn) for English language editing.
Footnotes
Supplemental Digital Content is available for this article. Direct URL citations are provided in the HTML and PDF versions of this article on the journal’s website, www.lww.com/international-journal-of-surgery.
Published online 24 January 2025
Contributor Information
Yongqi Dong, Email: dyq540736357@163.com.
Zijing Liu, Email: 457935001@qq.com.
Gang Zhao, Email: zhaogang0133@126.com.
Zhihuan Zhang, Email: 1207547128@qq.com.
Yuan Xia, Email: 739297458@qq.com.
Shiyong Xiao, Email: xiaoshiyong@stu.hebmu.edu.cn.
Zebin Lin, Email: l752838666@gmail.com.
Ethical approval
This study was implemented after approval by the medical ethics committee of Wushan County People’s Hospital of Chongqing (NO.2024-53).
Consent
Not applicable.
Sources of funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author’s contribution
Y.D.: conceptualization: lead; formal analysis: lead; methodology: lead; software: lead; validation: lead; writing – original draft: lead; writing – review & editing: equal; H.X.: data curation: supporting; methodology: supporting; writing – review & editing: supporting; W.Y.: data curation: supporting; supervision: supporting; validation: supporting; writing – review & editing: supporting; Z.L.: funding acquisition: lead; methodology: supporting; G.Z.: methodology: supporting; validation: supporting; visualization: supporting; writing – review & editing: supporting; Z.Z.: data curation: supporting; formal analysis: supporting; methodology: supporting; software: supporting; writing – review & editing: supporting; Y.X.: methodology: supporting; software: equal; writing – review & editing: supporting; S.X.: formal analysis: supporting; methodology: supporting; validation: supporting; Q.Y.: data curation: supporting; writing – review & editing: supporting; Z.L.: conceptualization: equal; supervision: equal; writing – review & editing: supporting.
Conflicts of interest disclosure
The authors declare that there are no conflicts of interest regarding the publication of this article.
Research registration unique identifying number (UIN)
PROSPERO (No. CRD42023458361).
Guarantor
None.
Provenance and peer review
None.
Presentation
Not applicable.
Data availability statement
No other datasets were generated during and/or analyzed during the current study. All the information is available with the manuscript.
Appendices
The search strategy, included studies, extracted data, methods, and other materials are available in Supplementary Digital Content, Appendix.
References
- [1].Sung H, Ferlay J, Siegel RL, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2021;71:209–49. [DOI] [PubMed] [Google Scholar]
- [2].Allemani C, Matsuda T, Di Carlo V, et al. Global surveillance of trends in cancer survival 2000-14 (CONCORD-3): analysis of individual records for 37 513 025 patients diagnosed with one of 18 cancers from 322 population-based registries in 71 countries. Lancet 2018;391:1023–75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Wang VS, Hornick JL, Sepulveda JA, Mauer R, Poneros JM. Low prevalence of submucosal invasive carcinoma at esophagectomy for high-grade dysplasia or intramucosal adenocarcinoma in Barrett’s esophagus: a 20-year experience. Gastrointest Endosc 2009;69:777–83. [DOI] [PubMed] [Google Scholar]
- [4].Abe S, Iyer PG, Oda I, Kanai N, Saito Y. Approaches for stricture prevention after esophageal endoscopic resection. Gastrointest Endosc 2017;86:779–91. [DOI] [PubMed] [Google Scholar]
- [5].Ishihara R, Arima M, Iizuka T, et al. Endoscopic submucosal dissection/endoscopic mucosal resection guidelines for esophageal cancer. Dig Endosc 2020;32:452–93. [DOI] [PubMed] [Google Scholar]
- [6].Libânio D, Pimentel-Nunes P, Bastiaansen B, et al. Endoscopic submucosal dissection techniques and technology: European Society of Gastrointestinal Endoscopy (ESGE) technical review. Endoscopy 2023;55:361–89. [DOI] [PubMed] [Google Scholar]
- [7].Yang J, Wang X, Li Y, et al. Efficacy and safety of steroid in the prevention of esophageal stricture after endoscopic submucosal dissection: a network meta-analysis. J Gastroenterol Hepatol 2019;34:985–95. [DOI] [PubMed] [Google Scholar]
- [8].Jia Y, Guo B, Zhang W, Li E, Zhang Q. Efficacy and safety of triamcinolone acetonide in the prevention of esophageal stricture after endoscopic submucosal dissection: a meta-analysis. Dis Esophagus 2022;35:doac039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Peng S, Huang S, Xia H, et al. The role of polyglycolic acid sheets in the management of post-endoscopic submucosal dissection gastrointestinal bleeding and esophageal stricture: a PRISMA compliant systematic review and meta-analysis. Medicine 2022;101:e29770. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Zhang BZ, Zhang Y, Wang YD, et al. Stent placement to prevent strictures after esophageal endoscopic submucosal dissection: a systematic review and meta-analysis. Dis Esophagus 2021;34:doab015. [DOI] [PubMed] [Google Scholar]
- [11].Caldwell DM, Ades AE, Higgins JP. Simultaneous comparison of multiple treatments: combining direct and indirect evidence. BMJ 2005;331:897–900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. Int J Surg 2021;88:105906. [DOI] [PubMed] [Google Scholar]
- [13].Shea BJ, Reeves BC, Wells G, et al. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomised or non-randomised studies of healthcare interventions, or both. BMJ 2017;358:j4008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Cumpston MS, McKenzie JE, Welch VA, Brennan SE. Strengthening systematic reviews in public health: guidance in the Cochrane Handbook for Systematic Reviews of Interventions, 2nd edition. J Public Health (Oxf) 2022;44:e588–e592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Salanti G, Del Giovane C, Chaimani A, Caldwell DM, Higgins JP. Evaluating the quality of evidence from a network meta-analysis. PloS One 2014;9:e99682. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Cumpston M, Li T, Page MJ, et al. Updated guidance for trusted systematic reviews: a new edition of the Cochrane handbook for systematic reviews of interventions. Cochrane Database Syst Rev 2019;10:Ed000142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ 1997;315:629–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Neupane B, Richer D, Bonner AJ, Kibret T, Beyene J. Network meta-analysis using R: a review of currently available automated packages. PloS One 2014;9:e115065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Brooks SP, Gelman A. General methods for monitoring convergence of iterative simulations. J Comput Graph Stat 1998;7:434–55. [Google Scholar]
- [20].Salanti G, Higgins JP, Ades AE, Ioannidis JP. Evaluation of networks of randomized trials. Stat Methods Med Res 2008;17:279–301. [DOI] [PubMed] [Google Scholar]
- [21].Bai AD, Komorowski AS, Ckl L, et al. Intention-to-treat analysis may be more conservative than per protocol analysis in antibiotic non-inferiority trials: a systematic review. BMC Med Res Methodol 2021;21:75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Melsen WG, Bootsma MC, Rovers MM, Bonten MJ. The effects of clinical and statistical heterogeneity on the predictive values of results from meta-analyses. Clin Microbiol Infect 2014;20:123–29. [DOI] [PubMed] [Google Scholar]
- [23].Chaimani A, Higgins JP, Mavridis D, Spyridonos P, Salanti G. Graphical tools for network meta-analysis in STATA. PloS One 2013;8:e76654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Dias S, Welton NJ, Caldwell DM, Ades AE. Checking consistency in mixed treatment comparison meta-analysis. Stat Med 2010;29:932–44. [DOI] [PubMed] [Google Scholar]
- [25].Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: an overview and tutorial. J Clin Epidemiol 2011;64:163–71. [DOI] [PubMed] [Google Scholar]
- [26].Yang F, Hu Y, Shi Z, et al. The occurrence and development mechanisms of esophageal stricture: state of the art review. J Transl Med 2024;22:123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Zhang Y, Yan X, Huang Y, et al. Efficacy of oral steroid gel in preventing esophageal stricture after extensive endoscopic submucosal dissection: a randomized controlled trial. Surg Endosc 2022;36:402–12. [DOI] [PubMed] [Google Scholar]
- [28].Zhang Y, Zhang B, Wang Y, et al. Advances in the prevention and treatment of esophageal stricture after endoscopic submucosal dissection of early esophageal cancer. J Transl Int Med 2020;8:135–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Hikichi T, Nakamura J, Takasumi M, et al. Prevention of stricture after endoscopic submucosal dissection for superficial esophageal cancer: a review of the literature. J Clin Med 2020;10:20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Ezoe Y, Muto M, Horimatsu T, et al. Efficacy of preventive endoscopic balloon dilation for esophageal stricture after endoscopic resection. J Clin Gastroenterol 2011;45:222–27. [DOI] [PubMed] [Google Scholar]
- [31].Tsujii Y, Hayashi Y, Kawai N, et al. Risk of perforation in balloon dilation associated with steroid injection for preventing esophageal stricture after endoscopic submucosal dissection. Endosc Int Open 2017;5:E573–579. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Yang YG, Liu CF, Hu JH, Lu GX. Polyglycolic acid sheet plus esophageal stent presents with higher efficacy in preventing esophageal stricture post endoscopic submucosal dissection compared with stent placement alone in early-stage esophageal cancer patients. Int J Clin Exp Med 2019;12:762–70. [Google Scholar]
- [33].Zhang Y, Wang X, Liu L, Chen JP, Fan ZN. Intramuscular injection of mitomycin C combined with endoscopic dilation for benign esophageal strictures. J Dig Dis 2015;16:370–76. [DOI] [PubMed] [Google Scholar]
- [34].Li L, Linghu E, Chai N, et al. Clinical experience of using a novel self-help inflatable balloon to prevent esophageal stricture after circumferential endoscopic submucosal dissection. Dig Endosc 2019;31:453–59. [DOI] [PubMed] [Google Scholar]
- [35].Saito Y, Tanaka T, Andoh A, et al. Usefulness of biodegradable stents constructed of poly-l-lactic acid monofilaments in patients with benign esophageal stenosis. World J Gastroenterol 2007;13:3977–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [36].Liao Z, Liao G, Yang X, et al. Transplantation of autologous esophageal mucosa to prevent stricture after circumferential endoscopic submucosal dissection of early esophageal cancer (with video). Gastrointest Endosc 2018;88:543–46. [DOI] [PubMed] [Google Scholar]
- [37].Ye S, Hu J, Zhang D, et al. Strategies for preventing esophageal stenosis after endoscopic submucosal dissection and progress in stem cell-based therapies. Tissue Eng Part B Rev 2024;30:522–29. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
No other datasets were generated during and/or analyzed during the current study. All the information is available with the manuscript.