Advertisement

Management of perforation related to endoscopic submucosal dissection for superficial duodenal epithelial tumors

Published:September 26, 2019DOI:https://doi.org/10.1016/j.gie.2019.09.024

      Background and Aims

      Endoscopic submucosal dissection (ESD) is being performed more frequently as a local treatment for superficial duodenal epithelial tumors (SDETs). However, ESD for SDETs is technically difficult because of specific anatomic features that increase the risk of perforation and often require surgery. This study was performed to evaluate the management of ESD-related perforation in patients with SDETs.

      Methods

      Patients who underwent ESD for SDETs from July 2010 to December 2018 were studied. We collected data on complete closure, insertion of endoscopic nasobiliary and pancreatic duct drainage (ENBPD) tubes, and additional interventions. We also evaluated clinical outcomes, including the fasting period, hospital stay, and maximum serum C-reactive protein level.

      Results

      ESD was completed in 264 patients with SDETs. Perforation was observed in 36 patients, including 4 patients with delayed perforation. Among 32 patients with intraoperative perforation, complete closure was achieved in 13 patients. Compared with patients without complete closure, the fasting period and hospital stay were significantly shorter and the maximum serum C-reactive protein level was significantly lower in patients with complete closure, which were equivalent to those in patients without perforation. In patients without complete closure for mucosal defect, no additional interventions were required when an ENBPD tube was inserted, whereas 2 patients without ENBPD tube insertion underwent additional interventions such as percutaneous drainage and a surgical operation.

      Conclusions

      Perforation associated with ESD for SDETs required complex conservative management with complete closure or insertion of an ENBPD tube.

      Abbreviations:

      CRP (C-reactive protein), ENBPD (endoscopic nasobiliary and pancreatic duct drainage), ESD (endoscopic submucosal dissection), NP (no perforation), P (perforation), PC (perforation with complete closure), PGA (polyglycolic acid), SDET (superficial duodenal epithelial tumor)
      The prevalence of superficial duodenal epithelial tumors (SDETs) is not as high as that of other organs of the digestive tract.
      • Jepsen J.M.
      • Persson M.
      • Jakobsen N.O.
      • et al.
      Prospective study of prevalence and endoscopic and histopathologic characteristics of duodenal polyps in patients submitted to upper endoscopy.
      ,
      • Murray M.A.
      • Zimmerman M.J.
      • Ee H.C.
      Sporadic duodenal adenoma is associated with colorectal neoplasia.
      However, the opportunity to detect SDETs has increased with recent advances in endoscopic technology.
      • Goda K.
      • Kikuchi D.
      • Yamamoto Y.
      • et al.
      Endoscopic diagnosis of superficial non-ampullary duodenal epithelial tumors in Japan: multicenter case series.
      Endoscopic resection is applied for SDETs as well as lesions in the esophagus, stomach, and colon
      • Yamamoto Y.
      • Yoshizawa N.
      • Tomida H.
      • et al.
      Therapeutic outcomes of endoscopic resection for superficial non-ampullary duodenal tumor.
      as an alternative to invasive surgical procedures such as pancreaticoduodenectomy. In particular, endoscopic submucosal dissection (ESD) achieves a higher en bloc resection rate than EMR.
      • Marques J.
      • Baldaque-Silva F.
      • Pereira P.
      • et al.
      Endoscopic mucosal resection and endoscopic submucosal dissection in the treatment of sporadic nonampullary duodenal adenomatous polyps.
      • Tsujii Y.
      • Nishida T.
      • Nishiyama O.
      • et al.
      Clinical outcomes of endoscopic submucosal dissection for superficial esophageal neoplasms: a multicenter retrospective cohort study.
      • Yamada T.
      • Hirota M.
      • Tsutsui S.
      • et al.
      Gastric endoscopic submucosal dissection under steady pressure automatically controlled endoscopy (SPACE): a multicenter randomized preclinical trial.
      • Kato M.
      • Nakajima K.
      • Yamada T.
      • et al.
      Esophageal submucosal dissection under steady pressure automatically controlled endoscopy (SPACE): a clinical feasibility study.
      • Hirota M.
      • Kato M.
      • Yamasaki M.
      • et al.
      A novel endoscopic submucosal dissection technique with robust and adjustable tissue traction.
      • Kato M.
      • Nishida T.
      • Yamamoto K.
      • et al.
      Scheduled endoscopic surveillance controls secondary cancer after curative endoscopic resection for early gastric cancer: a multicentre retrospective cohort study by Osaka University ESD study group.
      • Kato M.
      • Yahagi N.
      Advanced endoscopic treatment of gastric and duodenal neoplasms: beyond standard EMR and ESD.
      However, ESD for SDETs is technically difficult, with a reported intraoperative perforation rate ranging from 21.4% to 35.7%.
      • Matsumoto S.
      • Miyatani H.
      • Yoshida Y.
      Endoscopic submucosal dissection for duodenal tumors: a single-center experience.
      • Jung J.H.
      • Choi K.D.
      • Ahn J.Y.
      • et al.
      Endoscopic submucosal dissection for sessile, nonampullary duodenal adenomas.
      • Hoteya S.
      • Furuhata T.
      • Takahito T.
      • et al.
      Endoscopic submucosal dissection and endoscopic mucosal resection for non-ampullary superficial duodenal tumor.
      In contrast, the perforation rates in other digestive tract organs such as the esophagus, stomach, and colon range from 2.5% to 5.7%.
      • Takahashi H.
      • Arimura Y.
      • Masao H.
      • et al.
      Endoscopic submucosal dissection is superior to conventional endoscopic resection as a curative treatment for early squamous cell carcinoma of the esophagus (with video).
      • Fujiya M.
      • Tanaka K.
      • Dokoshi T.
      • et al.
      Efficacy and adverse events of EMR and endoscopic submucosal dissection for the treatment of colon neoplasms: a meta-analysis of studies comparing EMR and endoscopic submucosal dissection.
      • Hirao M.
      • Yamada T.
      • Michida T.
      • et al.
      Peritoneal seeding after gastric perforation during endoscopic submucosal dissection for gastric cancer.
      These high perforation rates are considered to be because of anatomic features such as the thin wall of the duodenal mucosa and limited maneuverability of the endoscope. Moreover, delayed perforation reportedly occurs in 6.30% of patients after ESD for SDETs,
      • Inoue T.
      • Uedo N.
      • Yamashina T.
      • et al.
      Delayed perforation: a hazardous complication of endoscopic resection for non-ampullary duodenal neoplasm.
      whereas this rate is only .40% to .45% in patients who undergo ESD for stomach and colon lesions.
      • Hanaoka N.
      • Uedo N.
      • Ishihara R.
      • et al.
      Clinical features and outcomes of delayed perforation after endoscopic submucosal dissection for early gastric cancer.
      ,
      • Saito Y.
      • Uraoka T.
      • Yamaguchi Y.
      • et al.
      A prospective, multicenter study of 1111 colorectal endoscopic submucosal dissections (with video).
      Although the precise mechanism of delayed perforation is unclear, bile and pancreatic juice containing digestive enzymes could play a pivotal role in the increased risk of delayed adverse events.
      • Hoteya S.
      • Kaise M.
      • Iizuka T.
      • et al.
      Delayed bleeding after endoscopic submucosal dissection for non-ampullary superficial duodenal neoplasias might be prevented by prophylactic endoscopic closure: analysis of risk factors.
      • Yahagi N.
      • Kato M.
      • Ochiai Y.
      • et al.
      Outcomes of endoscopic resection for superficial duodenal epithelial neoplasia.
      • Kato M.
      • Ochiai Y.
      • Fukuhara S.
      • et al.
      Clinical impact of closure of the mucosal defect after duodenal ESD.
      Duodenal ESD is reportedly associated with a high risk of intraoperative and delayed perforation. Nevertheless, little is known about the clinical course of patients with perforation after duodenal ESD because of the limited number of patients with this disease. Therefore, the present study was performed to elucidate the detailed clinical course and appropriate management of patients with perforation associated with ESD for SDETs.

      Methods

      Study design

      This retrospective case series was conducted in a university hospital. We identified consecutive patients with SDETs who underwent ESD from July 2010 to December 2018. Our database included all patients who underwent endoscopic treatment for duodenal lesions, and patients have been registered prospectively since February 2017. This study was performed in accordance with the 2008 revision of the Declaration of Helsinki. The study protocol was approved by our institutional review board (approval nos. 20150221 and 20160267).

      ESD procedure

      The ESD procedures were mainly performed using a therapeutic endoscope with a water jet function (GIF-Q260J; Olympus Medical Systems, Tokyo, Japan) under carbon dioxide insufflation. In some cases in which it was difficult to approach the lesion because of the distal location of the lesion or shape of the stomach (such as cascade stomach), a pediatric colonoscope (PCF-Q260JI; Olympus Medical Systems) was used. Based on our experience, we evaluated the risk factors of technical difficulty of duodenal ESD before the procedure, as recently reported.
      • Kato M.
      • Sasaki M.
      • Mizutani M.
      • et al.
      Predictors of technical difficulty with duodenal ESD.
      General anesthesia was then administered for high-difficulty cases, whereas benzodiazepine, pethidine hydrochloride, and dexmedetomidine were administered for low-difficulty cases. Submucosal injection was performed with 10% glycerine solution (Glyceol; Chugai Pharmaceutical Co, Ltd, Tokyo, Japan), and .4% sodium hyaluronate (MucoUp; Boston Scientific Japan, Tokyo, Japan) was used as required. Mucosal incision or submucosal dissection was performed using a 1.5-mm-long DualKnife or DualKnife J (Olympus Medical Systems). In difficult cases, such as those involving severe fibrosis or poor maneuverability, submucosal dissection was carefully performed using a HookKnife (Olympus Medical Systems). Minor bleeding was treated with these devices by placing the tip of the device into the outer sheath; however, when spurting bleeding occurred, a pair of hemostatic forceps (Coagrasper; Olympus Medical Systems) was used. These devices were powered by a high-frequency electrosurgical unit (VIO 300D; Erbe Elektromedizin, Tübingen, Germany) with dry cut (effect, 3-30 W) for mucosal incision, swift coagulation (effect, 4-30 W) for submucosal dissection, and soft coagulation (effect, 5-50 W) for hemostasis.

      Treatment strategy for perforation

      Intraoperative perforation during the ESD procedure was managed either by simple closure using clips (EZ Clip; Olympus Medical Systems) or by the string clip suturing technique
      • Yahagi N.
      • Nishizawa T.
      • Akimoto T.
      • et al.
      New endoscopic suturing method: string clip suturing method.
      to prevent delayed adverse events. Closure was performed after ESD for the entire mucosal defect, including the site of perforation. In some cases, when time was required to finish ESD after perforation, the perforation site was closed immediately to avoid leakage of duodenal contents. In addition, clipping was carried out to include the muscle layer so as not to form a space between the mucosa and muscle layer. If the wound was difficult to close completely, it was principally covered with polyglycolic acid (PGA) sheets and fibrin glue or regionally closed by a clip. Thereafter, we also attempted placement of an endoscopic nasobiliary and pancreatic duct drainage (ENBPD) tube in a subset of these cases to avoid exposure to bile and pancreatic juice, especially for lesions located around Vater’s papilla.
      Delayed perforation was defined as perforation manifested clinically, such as fever and abdominal pain, and revealed by imaging studies such as CT with fluid collection and abnormal air outside the duodenum, which did not occur during ESD. Urgent endoscopy was performed if possible to assess the status of the wound, to close or shield the perforated site, or to insert an ENBPD tube.

      ENBPD procedure

      As mentioned above, ENBPD was considered when complete closure of the post-ESD mucosal defect was impossible. The ENBPD procedure was mainly performed using a side-viewing endoscope (TJF-260V; Olympus Medical Systems). The side-viewing endoscope was carefully inserted to avoid injury to the duodenal wall, especially when advancing through the superior duodenal angle and stretching the endoscope. In patients who underwent selective cholangiography and pancreatography, a catheter (ERCP-Katheter, 0120211; MTW Endoskopie, Wesel, Germany) was then advanced into the common bile duct and main pancreatic duct with a .025-inch hydrophilic guidewire (VisiGlide 2; Olympus Medical Systems). If each deep cannulation was difficult, another .025-inch hydrophilic guide wire (NaviPro; Boston Scientific, Natick, Mass, USA) was applied. Finally, a 5F pigtail catheter (QuickPlace V ENBD, PBD-V813W-05; Olympus Medical Systems) was inserted into the common bile duct for endoscopic nasobiliary drainage, and a 5F straight catheter (Nasal Pancreatic Drainage Set, G21466; Cook Japan, Tokyo, Japan) was inserted into the main pancreatic duct for endoscopic nasopancreatic drainage. The endoscopic nasobiliary drainage and endoscopic nasopancreatic drainage tubes were advanced together and left at a suitable position. All procedures were performed by 2 proficient endoscopists (S.F. and E.I.) who have performed more than 2000 cases of ERCP.

      Post-ESD management

      After ESD, patients received intravenous hydration with administration of proton pump inhibitors (rabeprazole 20 mg/day, lansoprazole 30 mg/day, esomeprazole 20 mg/day, or vonoprazan 20 mg/day). Intravenous antibiotics were also administered to patients with perforation. Regardless of whether an intraoperative perforation was present, patients were basically fasted for 2 days after ESD and then began a liquid and soft solid diet at daily intervals. Thereafter, patients were generally discharged 5 days after ESD. If patients developed fever, abdominal pain, leukocytosis, or significant elevation of the C-reactive protein (CRP) level as indicated by blood examinations, food intake was postponed until after these abnormalities had improved. ENBPD tubes were removed after confirmation of no leakage by endoscopy, GI contrast radiography, or CT.

      Study population

      For evaluation of management and outcomes, the study population was classified into the following 4 subgroups to investigate the detailed clinical courses of patients with perforation: patients without perforation (no perforation [NP] group), patients with perforation and complete closure (perforation with complete closure [PC] group), patients with perforation but without complete closure (perforation [P] group), and patients with delayed perforation.

      Measured outcomes

      We collected data on patients’ background (age, sex, tumor size, resected size, location, final pathologic diagnosis, general anesthesia, and operation time) and clinical outcomes of duodenal ESD (maximum serum CRP level, fasting period, length of hospital stay, complete closure, insertion of ENBPD tube, and additional interventions such as percutaneous procedures and surgical operations). For patients who underwent ENBPD tube insertion, adverse events such as post-ERCP pancreatitis and perforation induced by insertion of the side-viewing endoscope were evaluated to assess the safety of the ENBPD procedure. We compared these clinical outcomes among the NP, PC, and P groups.

      Statistical analysis

      All continuous values are presented as mean ± standard deviation. The unpaired Student t test or one-way analysis of variance with Bonferroni correlation was used to evaluate the statistical significance of differences. The Fisher exact test or Pearson χ2 test was used to evaluate categorical data. Statistical significance was considered for 2-sided P < .05. All statistical analyses were performed using SPSS software version 25.0 for Mac (SPSS Japan, Tokyo, Japan).

      Results

      Overview of the study

      An overview of the study flow is presented in Figure 1. During the study period, ESD was performed in 278 patients, including 11 with non-neoplastic lesions. Among the 267 patients with SDETs, ESD was completed in 264 (98.9%). Perforation occurred in 36 patients (13.6%), including 4 with delayed perforation (1.5%). Among the 32 patients with intraoperative perforation, complete closure was achieved in 13 (40.6%), whereas the perforations could not be completely closed in the remaining 19.
      Figure thumbnail gr1
      Figure 1Overview and flow of the study. Among 267 patients with superficial duodenal epithelial tumors, ESD was completed in 264 patients. Perforation occurred in 36 patients, including 4 with delayed perforation. Among 32 patients with intraoperative perforation, complete closure with endoscopic clipping was achieved in 13, whereas 19 patients had incomplete closure or underwent no procedure related to closure. ESD, Endoscopic submucosal dissection.

      Patients’ clinical features

      Dividing all 264 cases with ESD for SDETs into consecutive thirds, the perforations associated with duodenal ESD occurred in 15 (17.0%), 14 (15.9%), and 7 (8.0%) in each consecutive 88 cases, respectively. However, there was no statistically significant trend (Fig. 2).
      Figure thumbnail gr2
      Figure 2Changes of perforation rate in cases with duodenal endoscopic submucosal dissection (ESD). The rate of perforation associated with ESD for superficial duodenal epithelial tumors (n = 264) was calculated by dividing the total into 3 consecutive cohorts of 88 cases. Compared with that in the first 88 cases, it tended to decrease in the third 88 cases, although without statistical significance in trend (P = .079). Statistical analysis was performed by the Pearson χ2 test.
      Table 1 shows the patients’ background features. Their mean age and tumor size were 62.8 ± 11.8 years and 29.8 ± 17.6 mm, respectively. Thirty-nine lesions were located on the duodenal bulb, and the other 225 lesions were located distal to the superior duodenal angle. The pathologic findings showed adenocarcinoma in 91 cases (34.5%). The tumor size and resected size were significantly larger in patients with than without perforation. Moreover, the operation time was significantly longer in patients with than without perforation. These findings might reflect the complicated situation of ESD with perforation. R0 resection was achieved in 219 of 264 cases (83.0%).
      Table 1Background of patients who underwent complete duodenal ESD procedures
      Without perforation (n = 228)With perforation (n = 36)P value
      Age, y62.7 ± 11.963.2 ± 11.4.815
      Sex, male/female152/7624/121.000
      Size of lesion, mm28.6 ± 17.237.6 ± 18.2.004
      Resected size, mm37.9 ± 18.047.2 ± 17.7.004
      Location (bulb/distal to SDA)36/1923/33.317
      Pathology (adenoma/adenocarcinoma)158/7015/21.249
      General anesthesia, %103 (45.2)21 (58.3).154
      Operation time, min59.4 ± 43.887.3 ± 67.3.001
      Values are mean ± standard deviation or n (%).
      ESD, Endoscopic submucosal dissection; SDA, superior duodenal angle.

      Outcomes of patients with perforation according to the degree of closure

      The outcomes of complete closure of mucosal defects in patients with intraoperative perforation are shown in Figure 3. The mean fasting period in the P group (9.35 ± 1.74 days) was significantly longer than that in the PC and NP groups (3.42 ± .34 and 2.36 ± .67 days, respectively; P < .001) (Fig. 3A). There was no significant difference between the PC and NP groups. Similarly, the length of hospital stay in the P group (21.18 ± .64 days) was significantly longer than that in the PC and NP groups (7.17 ± .89 and 5.14 ± .15 days, respectively; P < .001) (Fig. 3B), whereas it was not significantly different between the PC and NP groups. Moreover, the maximum serum CRP level in the P group (11.67 ± 2.27 mg/dL) was significantly higher than that in the PC and NP groups (3.68 ± .82 and 1.72 ± .21 mg/dL, respectively; P < .001) (Fig. 3C), whereas there was no significant difference between the PC and NP groups. Thus, even if intraoperative perforation occurred by ESD for an SDET, complete closure could allow for a good clinical course, even in patients without perforation.
      Figure thumbnail gr3
      Figure 3Comparison of outcomes among patients without perforation (NP group), with perforation and complete closure (PC group), and without complete closure (P group). A, The average fasting period was significantly longer in the P group (black bar) than in the NP group (white bar) and PC group (gray bar). In contrast, it was not different between the NP and PC groups. B, The average hospital stay was significantly longer in the P group than in the NP and PC groups. In contrast, it was not different between the NP and PC groups. C, The average serum C-reactive protein level was significantly higher in the P group than in the NP and PC groups. In contrast, it was not different between the NP and PC groups. ***P < .001. All statistical analyses were performed by 1-way analysis of variance (Bonferroni method).

      Details of patients with incomplete closure of intraoperative perforation by duodenal ESD

      Because the clinical outcomes were significantly worse in patients with perforation and incomplete wound closure, we further investigated the clinical courses of these patients. Table 2 shows the details of each patient without complete closure of the perforation. Of these 19 patients, 3 in whom the lesions were located on the duodenal bulb had a good clinical course (short hospital stay and no need for additional interventions or insertion of an ENBPD tube). Of the 16 patients in whom the lesions were located distal to the superior duodenal angle, 8 (50.0%) underwent ENBPD tube insertion. These patients required no additional interventions and experienced no adverse events related to ERCP, such as post-ERCP pancreatitis. However, of the remaining 8 patients with neither complete closure nor ENBPD tube insertion, percutaneous drainage was required in 1 (12.5%) for treatment of an extended abdominal abscess.
      Table 2Details of patients treated by incomplete or no closure for mucosal defect with perforation during duodenal ESD
      Age (y)SexLocationCircumferenceSize (mm)Involvement of papillaENBPDAdditional interventionMaximum CRP (mg/dL)Hospital stay (days)
      63MBulbPosterior40NNN.14
      66MBulbPosterior15NNN.14
      73FSDAMedial60NNN3.74
      70MBulbPosterior25NNN1.26
      68MIDALateral67NNN2.06
      67MDescendingMedial28NNN3.57
      77FDescendingMedial46NYN.9610
      74MDescendingMedial20NNN14.310
      49MDescendingMedial52YYN10.7811
      66FDescendingLateral45NYN8.412
      55MIDALateral40NYN18.515
      73FDescendingPosterior55NNN11.217
      47MDescendingMedial12NNN17.125
      71FDescendingLateral45NNN18.826
      47FDescendingAnterior64YYN25.827
      66FDescendingAnterior50YYN4.638
      69MSDAMedial70NNPercutaneous drainage for abscess22.246
      56FDescendingLateral40NYN21.851
      70FDescendingMedial67NYN25.458
      ESD, Endoscopic submucosal dissection; ENBPD, endoscopic nasobiliary and pancreatic duct drainage; CRP, C-reactive protein; IDA, inferior duodenal angle; SDA, superior duodenal angle.

      Features of patients with delayed perforation after duodenal ESD

      We identified 4 patients (1.5%) with delayed perforation in the study population. The details of each patient are presented in Table 3. Delayed perforation was recognized 8 to 33 hours after ESD. All lesions were located at the descending part of the duodenum, and Vater’s papilla was involved in 1 patient. An ENBPD tube was inserted in 2 patients (Fig. 4), whereas 1 other patient without ENBPD tube insertion developed an extended abdominal abscess requiring pancreatoduodenectomy. The maximum serum CRP level was significantly higher in patients with delayed perforation than in those with intraoperative perforation (P = .001), suggesting that delayed perforation results in more severe inflammation and should be managed by appropriate treatment as soon as possible.
      Table 3Details of patients with delayed perforation after duodenal ESD
      Age (y)SexLocationCircumferenceSize (mm)Involvement of papillaOnset after ESD (h)ENBPD after onsetAdditional interventionMaximum CRP (mg/dL)Hospital stay (days)
      50MDescendingPosterior55N33NN29.99
      56MDescendingPosterior22N24YN16.514
      50MDescendingMedial28N8NSurgical operation for abscess26.739
      83MDescendingMedial12Y10YN43.042
      ESD, Endoscopic submucosal dissection; ENBPD, endoscopic nasobiliary and pancreatic duct drainage; CRP, C-reactive protein.
      Figure thumbnail gr4
      Figure 4Images of a patient with delayed perforation after duodenal endoscopic submucosal dissection (ESD). A, A whitish, flat, elevated lesion (Paris classification, 0-IIa) was detected in a 56-year-old man with a history of total colectomy and endoscopic papillectomy because of familial adenomatous polyposis. The size of the lesion was 22 mm, and it was located on the posterior wall of the descending part of the duodenum near Vater’s papilla. The lesion was resected by ESD in a single piece. B, Because of development of high-grade fever 24 hours after ESD a CT was taken, which showed abnormal air and fluid collection outside the duodenum, suggesting delayed perforation (white arrow). C, An urgent endoscopic examination was performed. Effusion of pus was observed from the space between the endoscopic clips. D, Closure of the perforation by endoscopic clipping was impossible because of the fragile tissue; therefore, the mucosal defect including the perforated site was covered with a polyglycolic acid sheet and fibrin glue. E, A 5F endoscopic nasobiliary and pancreatic duct drainage (ENBPD) tube was then inserted into the common bile duct and main pancreatic duct by a side-viewing endoscope. F, The clinical course of this patient was good. Although the serum CRP level (solid line) and body temperature (dotted line) were elevated after duodenal ESD, they dramatically improved after insertion of the ENBPD tube, and the patient was discharged 2 weeks after ESD with no further interventions. POD, Postoperative day; CRP, C-reactive protein; BT, body temperature.

      Discussion

      In the present retrospective study, we analyzed 264 patients who underwent ESD for SDETs, including 32 with intraoperative perforation and 4 with delayed perforation. To the best of our knowledge, this is the first study to focus on the detailed clinical outcomes of perforation associated with ESD for SDETs in such a high number of patients. This study revealed that complete closure of the whole mucosal defect, not only the perforation site, suppressed the inflammatory reaction and significantly shortened both the duration of fasting and hospital stay, and these patients’ clinical courses were comparable with those of patients without perforation. Furthermore, when complete closure of the entire mucosal defect is impossible, combination of shielding the wound by a PGA sheet and insertion of an ENBPD tube could be an alternative treatment to avoid additional interventions such as surgical operations or percutaneous drainage.
      Although ESD has been widely accepted as the standard treatment for epithelial neoplasia arising from various organs, duodenal ESD is still controversial and considered to be very high risk. In particular, perforation is one of the most dangerous adverse events, and surgery is required for 14.3% to 33.0% of perforations.
      • Jung J.H.
      • Choi K.D.
      • Ahn J.Y.
      • et al.
      Endoscopic submucosal dissection for sessile, nonampullary duodenal adenomas.
      ,
      • Hara Y.
      • Goda K.
      • Dobashi A.
      • et al.
      Short- and long-term outcomes of endoscopically treated superficial non-ampullary duodenal epithelial tumors.
      ,
      • Perez-Cuadrado-Robles E.
      • Queneherve L.
      • Margos W.
      • et al.
      ESD versus EMR in non-ampullary superficial duodenal tumors: a systematic review and meta-analysis.
      The duodenum is located in the retroperitoneum and is very close to the pancreas, bile duct, and other organs; therefore, the extension of inflammation is often asymptomatic and not easily noticed, retroperitoneal abscesses readily form, percutaneous drainage is difficult and has limited effectiveness, and surgical interventions tend to be invasive.
      • Inoue T.
      • Uedo N.
      • Yamashina T.
      • et al.
      Delayed perforation: a hazardous complication of endoscopic resection for non-ampullary duodenal neoplasm.
      However, ESD for SDETs can achieve en bloc resection, especially for large lesions,
      • Yahagi N.
      • Kato M.
      • Ochiai Y.
      • et al.
      Outcomes of endoscopic resection for superficial duodenal epithelial neoplasia.
      leading to a more accurate histologic diagnosis. According to a previous report, recurrence after endoscopic piecemeal mucosal resection for lateral spreading lesions of Vater’s papilla was more frequent compared with resection in a single piece.
      • Klein A.
      • Qi Z.
      • Bahin F.F.
      • et al.
      Outcomes after endoscopic resection of large laterally spreading lesions of the papilla and conventional ampullary adenomas are equivalent.
      Thus, ESD for SDETs is important for histologic evaluation and management after endoscopic resection, especially when the SDET contains a malignant lesion. Moreover, the rate of perforation tended to decrease in recent cases compared with those in the early part of the study period, suggesting that ESD for SDETs was becoming a safer procedure. One of the reasons for this may be the advantage gained by the use of several new methods such as application of water pressure, which makes cutting of the submucosal layer easy because of the improved visibility during ESD.
      • Yahagi N.
      • Nishizawa T.
      • Sasaki M.
      • et al.
      Water pressure method for duodenal endoscopic submucosal dissection.
      Because duodenal ESD still carries a high risk of adverse events and requires advanced techniques, development of new devices or methods will reduce these risks and provide further advantages for the treatment of SDET.
      In the present study, complete closure of the entire mucosal defect resulted in better clinical outcomes in terms of the fasting period, hospital stay, and serum CRP level than did incomplete closure or no procedure related to closure. Generally, closure of only the perforation site is adequate for intraoperative perforations in the esophagus, stomach, and colon.
      • Yamamoto Y.
      • Kikuchi D.
      • Nagami Y.
      • et al.
      Management of adverse events related to endoscopic resection of upper gastrointestinal neoplasms: review of the literature and recommendations from experts.
      There are 2 possible reasons for this difference between the duodenum and other organs. Unlike other organs of the digestive tract, the proper muscle layer is very thin in the duodenum and is easily torn by the tip of clips when trying to close the perforation site. Moreover, as mentioned above, exposure of the tissue to bile and pancreatic juice could aggravate the mucosal defect if complete closure is not achieved. Thus, the particular environment of the duodenum might require complete closure of the entire mucosal defect, including the perforated site, to improve outcomes.
      Although complete closure is important in managing perforation associated with duodenal ESD, it is sometimes difficult or even impossible depending on the location of the lesion. The clinical outcomes of lesions located at the duodenal bulb were good, and we were able to observe such patients without any particular intervention. One potential explanation for these good outcomes is that the bile and pancreatic juice do not directly affect the duodenal bulb, resulting in good outcomes even without complete closure of the mucosal defect or insertion of an ENBPD tube. For lesions located at the distal duodenum, shielding the wound with a PGA sheet is reportedly effective in avoiding delayed adverse events when complete closure of the mucosal defect is impossible.
      • Takimoto K.
      • Imai Y.
      • Matsuyama K.
      Endoscopic tissue shielding method with polyglycolic acid sheets and fibrin glue to prevent delayed perforation after duodenal endoscopic submucosal dissection.
      Nevertheless, in our study, patient outcomes were relatively worse when complete closure was not achieved, despite the fact that we used PGA sheets for shielding the defect in all those patients. One reason for these poorer outcomes could be that although PGA shielding is effective for prevention, it is not adequate in patients with perforation, which is a more serious situation, because the PGA sheet might become compromised by bile, pancreatic juice, or peristalsis. Further improvement in the management of these patients is needed.
      Unexpectedly, 3 of 4 patients with delayed perforation were successfully treated conservatively and required neither surgery nor percutaneous drainage. Delayed perforation is a rare but serious adverse event in the stomach. Hanaoka et al
      • Hanaoka N.
      • Uedo N.
      • Ishihara R.
      • et al.
      Clinical features and outcomes of delayed perforation after endoscopic submucosal dissection for early gastric cancer.
      reported 6 cases of delayed perforation after gastric ESD. The incidence rate was .45% (6/1329), and 5 of 6 patients (83.3%) required emergency surgery. In the present study, most patients were successfully treated by a multidisciplinary approach involving a combination of multiple clipping, filling with PGA sheets and fibrin glue, and ENBPD tube insertion. The hospital stay was relatively long, ranging from 9 to 42 days in these patients; however, this is not unexpected considering the invasiveness of pancreaticoduodenectomy. Remarkably, both patients who underwent ENBPD tube insertion were successfully treated by conservative therapy alone, whereas 1 of 2 patients required pancreaticoduodenectomy. External drainage of bile and pancreatic juice might be effective in improving the outcomes of patients with delayed perforation; however, the safety of inserting a side-viewing endoscope in patients with a large mucosal defect created by ESD and the incidence rate of post-ERCP pancreatitis should be more closely examined in further studies.
      This study has several limitations because of its single-center, retrospective design. All procedures were performed by experts in a high-volume center, preventing generalization of our results. The number of patients was limited, and there were no control subjects such as patients with large resected lesions who underwent neither endoscopic closure nor ENBPD because of the high risk associated with not performing any procedures for such patients. Furthermore, we could not eliminate a certain degree of selection bias. There was no standardized management protocol, and we relied on the clinicians’ subjective judgment regarding decisions such as the indication for placement of and timing of removal of the ENBPD tube. Because of these limitations, the results of our study should be interpreted carefully.
      In conclusion, the present study has shown the effectiveness of complete closure of mucosal defects with perforations during ESD for SDETs, especially those located at the distal duodenum. A multidisciplinary approach including ENBPD tube insertion might enable conservative treatment even in patients with delayed perforation.

      Acknowledgments

      We thank Angela Morben, DVM, ELS, from Edanz Group (www.edanzediting.com/ac), for editing a draft of this manuscript. We also appreciate Mari Mizutani, Marin Noda, Yujiro Machida, Hiroki Tamagawa, Yoichi Takimoto, Tadashi Katayama, Shintaro Kawasaki, Takashi Seino, Masayasu Horibe, and Naoki Hosoe for their contribution to patient care.

      References

        • Jepsen J.M.
        • Persson M.
        • Jakobsen N.O.
        • et al.
        Prospective study of prevalence and endoscopic and histopathologic characteristics of duodenal polyps in patients submitted to upper endoscopy.
        Scand J Gastroenterol. 1994; 29: 483-487
        • Murray M.A.
        • Zimmerman M.J.
        • Ee H.C.
        Sporadic duodenal adenoma is associated with colorectal neoplasia.
        Gut. 2004; 53: 261-265
        • Goda K.
        • Kikuchi D.
        • Yamamoto Y.
        • et al.
        Endoscopic diagnosis of superficial non-ampullary duodenal epithelial tumors in Japan: multicenter case series.
        Dig Endosc. 2014; 26: 23-29
        • Yamamoto Y.
        • Yoshizawa N.
        • Tomida H.
        • et al.
        Therapeutic outcomes of endoscopic resection for superficial non-ampullary duodenal tumor.
        Dig Endosc. 2014; 26: 50-56
        • Marques J.
        • Baldaque-Silva F.
        • Pereira P.
        • et al.
        Endoscopic mucosal resection and endoscopic submucosal dissection in the treatment of sporadic nonampullary duodenal adenomatous polyps.
        World J Gastrointest Endosc. 2015; 7: 720-727
        • Tsujii Y.
        • Nishida T.
        • Nishiyama O.
        • et al.
        Clinical outcomes of endoscopic submucosal dissection for superficial esophageal neoplasms: a multicenter retrospective cohort study.
        Endoscopy. 2015; 47: 775-783
        • Yamada T.
        • Hirota M.
        • Tsutsui S.
        • et al.
        Gastric endoscopic submucosal dissection under steady pressure automatically controlled endoscopy (SPACE): a multicenter randomized preclinical trial.
        Surg Endosc. 2015; 29: 2748-2755
        • Kato M.
        • Nakajima K.
        • Yamada T.
        • et al.
        Esophageal submucosal dissection under steady pressure automatically controlled endoscopy (SPACE): a clinical feasibility study.
        Endoscopy. 2014; 46: 680-684
        • Hirota M.
        • Kato M.
        • Yamasaki M.
        • et al.
        A novel endoscopic submucosal dissection technique with robust and adjustable tissue traction.
        Endoscopy. 2014; 46: 499-502
        • Kato M.
        • Nishida T.
        • Yamamoto K.
        • et al.
        Scheduled endoscopic surveillance controls secondary cancer after curative endoscopic resection for early gastric cancer: a multicentre retrospective cohort study by Osaka University ESD study group.
        Gut. 2013; 62: 1425-1432
        • Kato M.
        • Yahagi N.
        Advanced endoscopic treatment of gastric and duodenal neoplasms: beyond standard EMR and ESD.
        Am J Gastroenterol. 2018; 113: 1423-1426
        • Matsumoto S.
        • Miyatani H.
        • Yoshida Y.
        Endoscopic submucosal dissection for duodenal tumors: a single-center experience.
        Endoscopy. 2013; 45: 136-137
        • Jung J.H.
        • Choi K.D.
        • Ahn J.Y.
        • et al.
        Endoscopic submucosal dissection for sessile, nonampullary duodenal adenomas.
        Endoscopy. 2013; 45: 133-135
        • Hoteya S.
        • Furuhata T.
        • Takahito T.
        • et al.
        Endoscopic submucosal dissection and endoscopic mucosal resection for non-ampullary superficial duodenal tumor.
        Digestion. 2017; 95: 36-42
        • Takahashi H.
        • Arimura Y.
        • Masao H.
        • et al.
        Endoscopic submucosal dissection is superior to conventional endoscopic resection as a curative treatment for early squamous cell carcinoma of the esophagus (with video).
        Gastrointest Endosc. 2010; 72: 255-264
        • Fujiya M.
        • Tanaka K.
        • Dokoshi T.
        • et al.
        Efficacy and adverse events of EMR and endoscopic submucosal dissection for the treatment of colon neoplasms: a meta-analysis of studies comparing EMR and endoscopic submucosal dissection.
        Gastrointest Endosc. 2015; 81: 583-595
        • Hirao M.
        • Yamada T.
        • Michida T.
        • et al.
        Peritoneal seeding after gastric perforation during endoscopic submucosal dissection for gastric cancer.
        Dig Surg. 2018; 35: 457-460
        • Inoue T.
        • Uedo N.
        • Yamashina T.
        • et al.
        Delayed perforation: a hazardous complication of endoscopic resection for non-ampullary duodenal neoplasm.
        Dig Endosc. 2014; 26: 220-227
        • Hanaoka N.
        • Uedo N.
        • Ishihara R.
        • et al.
        Clinical features and outcomes of delayed perforation after endoscopic submucosal dissection for early gastric cancer.
        Endoscopy. 2010; 42: 1112-1115
        • 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
        • Hoteya S.
        • Kaise M.
        • Iizuka T.
        • et al.
        Delayed bleeding after endoscopic submucosal dissection for non-ampullary superficial duodenal neoplasias might be prevented by prophylactic endoscopic closure: analysis of risk factors.
        Dig Endosc. 2015; 27: 323-330
        • Yahagi N.
        • Kato M.
        • Ochiai Y.
        • et al.
        Outcomes of endoscopic resection for superficial duodenal epithelial neoplasia.
        Gastrointest Endosc. 2018; 88: 676-682
        • Kato M.
        • Ochiai Y.
        • Fukuhara S.
        • et al.
        Clinical impact of closure of the mucosal defect after duodenal ESD.
        Gastrointest Endosc. 2018; 89: 87-93
        • Kato M.
        • Sasaki M.
        • Mizutani M.
        • et al.
        Predictors of technical difficulty with duodenal ESD.
        Endosc Int Open. 2019; 6 (e1204-5)
        • Yahagi N.
        • Nishizawa T.
        • Akimoto T.
        • et al.
        New endoscopic suturing method: string clip suturing method.
        Gastrointest Endosc. 2016; 84: 1064-1065
        • Hara Y.
        • Goda K.
        • Dobashi A.
        • et al.
        Short- and long-term outcomes of endoscopically treated superficial non-ampullary duodenal epithelial tumors.
        World J Gastroenterol. 2019; 25: 707-718
        • Perez-Cuadrado-Robles E.
        • Queneherve L.
        • Margos W.
        • et al.
        ESD versus EMR in non-ampullary superficial duodenal tumors: a systematic review and meta-analysis.
        Endosc Int Open. 2018; 6: E998-E1007
        • Klein A.
        • Qi Z.
        • Bahin F.F.
        • et al.
        Outcomes after endoscopic resection of large laterally spreading lesions of the papilla and conventional ampullary adenomas are equivalent.
        Endoscopy. 2018; 50: 972-983
        • Yahagi N.
        • Nishizawa T.
        • Sasaki M.
        • et al.
        Water pressure method for duodenal endoscopic submucosal dissection.
        Endoscopy. 2017; 49: E227-E228
        • Yamamoto Y.
        • Kikuchi D.
        • Nagami Y.
        • et al.
        Management of adverse events related to endoscopic resection of upper gastrointestinal neoplasms: review of the literature and recommendations from experts.
        Dig Endosc. 2019; 31: 4-20
        • Takimoto K.
        • Imai Y.
        • Matsuyama K.
        Endoscopic tissue shielding method with polyglycolic acid sheets and fibrin glue to prevent delayed perforation after duodenal endoscopic submucosal dissection.
        Dig Endosc. 2014; 26: 46-49

      Linked Article

      • Duodenal endoscopic submucosal dissection: Is it ready for primetime? (with video)
        Gastrointestinal EndoscopyVol. 91Issue 5
        • Preview
          Endoscopic submucosal dissection (ESD) was developed in Japan and is now widely used around the world.1-3 Compared with piecemeal EMR, ESD allows en bloc removal of GI tract lesions, decreasing the rate of postprocedural recurrence and allowing adequate histologic interpretation of the resected specimen.4-6 However, ESD remains a meticulous, time-consuming, labor-intensive, and difficult-to-learn technique.4
        • Full-Text
        • PDF