Abstract
Background:
The optimal operative process of esophageal endoscopic submucosal dissection (ESD), especially for the beginners, is not established. In this study, the clip-with-line method, the tunneling method and the conventional method for esophageal ESD were compared in a training course on live pigs.
Methods:
15 trainee endoscopists were randomized into three groups, the clip-with-line method group, the tunneling method group, and the conventional method group. Each trainee performed four esophageal ESDs on live pigs with one of the specified methods, assisted by a senior endoscopist. The procedural time and speed, the en bloc resection rates, adverse events, and self-completion rates were recorded, and learning curves were drawn.
Results:
The procedural time in the clip-with-line group was significantly shorter than those in the tunneling and the conventional method group (47.4 ± 9.0 min versus 67.0 ± 15.1 min and 67.0 ± 11.8 min, p = 0.000). The clip-with-line method had the lowest rates of perforation and muscle layer injury among the three methods. The en bloc resection rates and self-completion rates were similar among the three groups. Learning curves analysis showed the clip-with-line method was the easiest one for the trainees to master.
Conclusions:
The clip-with-line method shortened procedural time for trainees, and was associated with lower rates of adverse events. This method is preferred for endoscopists in their learning periods for esophageal ESD.
Keywords: animal model, endoscopic submucosal dissection, esophagus, traction method, training
Introduction
Endoscopic submucosal dissection (ESD) is a technique that is developed for removal of superficial gastrointestinal tumors including early cancers or high-grade dysplasia. It achieves higher en bloc and curative resection rates of large lesions as well as reduces the risk of local recurrence compared with endoscopic mucosal resection (EMR) [Cao et al. 2009]. However, ESD is associated with more adverse events such as bleeding and perforation than EMR, and requires considerable expertise and prolonged operation time [Cao et al. 2009; Teoh et al. 2010; Rosch et al. 2004]. Esophageal ESD is technically more difficult than gastric ESD due to the narrower space of esophageal lumen [Fujishiro et al. 2009]. Also, the risk of perforation is higher because of the thin muscle layer of the esophageal wall [Maeda et al. 2011]. Thus, it is suggested that only operators who have adequate skill should perform esophageal ESD [Oyama, 2014].
Various techniques have been developed in order to overcome the technical difficulties of esophageal ESD, such as new endoknives and high-viscosity solutions for submucosal injection [Oyama, 2014]. It was reported that traction using a clip-with-line method was useful for enhancing the visibility of the operation field during ESD [Oyama et al. 2002]. Several retrospectively clinical case-control studies demonstrated that the method shortened operation time as well as ensured the safety of ESD [Suzuki et al. 2016; Ota et al. 2012]. A tunneling method, or called endoscopic submucosal tunnel dissection, was also reported to facilitate removal of esophageal lesions, especially large ones [Oyama et al. 2006; Linghu et al. 2013]. However, the optimal operative process of esophageal ESD, especially for the beginners, is not yet established.
Training on animal models could help trainees shorten the initial learning curve before starting ESD in humans. Therefore, in the current study, we compared the clip-with-line method with the tunneling method and the conventional esophageal ESD in a training course on live pigs.
Methods
ESD training program
The study was carried out in the Olympus China Medical Training and Education Center (Shijingshan, Beijing, China) between January 2015 and August 2015. A total of 15 trainees were selected to participate in this study who met the following criteria: with experience of performed more than 5000 upper gastrointestinal endoscopies, more than 200 EMRs or polypectomy, at least 1 but no more than 5 gastric or rectal ESDs, without experience of performing esophageal ESD and ESD on animal models. These 15 trainees were randomized into three groups (the clip-with-line method group, the tunneling method group, and the conventional method group), with 5 in each group. There was a senior endoscopist in each group to assist the trainees in the ESD procedures. Each senior endoscopist had experience of performing more than 500 ESDs, including at least 100 esophageal ESDs.
In each group, a video of esophageal ESD on real clinical cases with the corresponding method (not the other two methods) was shown to the trainees before performing ESD on the live pigs. Then the trainees observed esophageal ESDs performed by the senior endoscopists on live pigs with the specific method, and participated as assistants. After that, each trainee performed four esophageal ESDs on live pigs, assisted by a senior endoscopist with verbal advice, mainly to avoid adverse events and to ensure that the correct operative process of ESD was carried out. Situations when the procedures were taken over by senior endoscopists included: (1) severe complications, such as large perforation and unstopped bleeding or (2) unable to continue dissection. In these cases, the adverse events would be managed by senior endoscopists, and then the trainees continued the procedures when the events were resolved. For each trainee, the entire training course lasted for two days.
Animal models
Pigs were fasted for 24 h prior to the procedures. ESD was carried out under general anesthesia using intravenous injection of propofol with tracheal intubation. A piece of mucosa on the posterior wall at the middle or lower part of the porcine esophagus was chosen to be a simulated lesion in each procedure. The posterior wall of the esophagus was determined by the following method: with the pig laying on its left side, the liquid in the esophageal lumen should be accumulated in the 9 o’clock area, then the posterior wall was located at the 6 o’clock direction. Marking spots were placed every 5 mm around a simulated lesion. The size of the marked area was same in each group: 30mm in longitudinal diameter, and about a quarter of the esophageal circumference. Each pig underwent 2–3 esophageal ESDs before being humanely euthanized. The intervals between two adjacent simulated lesions on the same pig were at least 20 mm, in order to avoid disturbing the next procedure.
The study was approved by our Institutional Review Board and Animal Ethics Committee (BZY2015005), and animals were managed in accordance with Chinese laws for animal use and care.
ESD
A single-channel video endoscope (GIF-Q260J, Olympus, Tokyo, Japan) with a transparent hood (D-201-11804, Olympus) attached at its top was used for all procedures. Submucosal injections of normal saline solution mixed with a small amount of indigo carmine were administered for the elevation of the mucosa and submucosa. The added indigo carmine facilitated the identification of the space between the muscularis propria and submucosa by creating a light blue color. An ESG-100 electrosurgical generator (Olympus) was used as a power source for all ESD procedures. The knives and the power generator settings for each step were as follows: marking, dual knives (KD-650Q, Olympus), soft-coagulation mode 80W; marginal incision, dual knives or IT2 knives (KD-611L, Olympus), pulse-cut-slow mode 30W, and submucosal dissection or hemostasis, dual knives or IT2 knives, forced-coagulation-1 mode 30W. The setting for a coagrasper (FD-410LR, Olympus) was soft-coagulation mode 80W for hemostasis.
The clip-with-line method (Figure 1): A ‘clip-with-line’ was assembled by attaching a nylon string (e.g. dental floss) to a standard endoscopic clip (HX-610-090, Olympus) (Figure 2). Mucosal incision was performed on the anal side by a dual knife, then circumferential incision was accomplished by an IT2 knife. After circumferential incision, adequate submucosal dissection was obtained to expose enough tissue to be grasped later by an endoclip. The endoscope was then removed. A hemoclip applicator device (HX-110LR, Olympus) was inserted into the accessory channel of the endoscope, and an endoclip was mounted onto the tip of the clip device. Then, dental floss was tied to one arm of the clip by a surgeon’s knot. Before intubation of the esophagus, the endoclip was withdrawn inside the channel to avoid trauma during insertion of the endoscope. When the scope reached the lesion, the clip was tied with a dental floss and attached to the oral edge of the lesion, and the line was pulled gently in the oral direction by the endoscopist or an assistant for traction. Then submucosal dissection was completed by an IT2 knife.
Figure 1.
Esophageal ESD using the clip-with-line method on live pigs.
(a) The operative process was as follows: (i): Mucosal incision was performed on the anal side. (ii) & (iii): Circumferential incision was carried out. (iv): The clip-with-line was attached to the oral edge for traction. (v): Submucosal dissection was carried out.
(b) Mucosal incision was performed firstly on the anal side by a dual knife.
(c) & (d) Circumferential incision was completed by an IT2 knife.
(e) The clip tied by a dental floss was attached to the oral edge of the lesion, and the line was pulled in the oral direction for traction.
(f) Submucosal dissection was carried out by an IT2 knife.
ESD, endoscopic submucosal dissection.
Figure 2.
A ‘clip-with-line’ was assembled by attaching a nylon string to a standard endoscopic clip, which was mounted onto the tip of a hemoclip applicator device inserted into the accessory channel of an endoscope (a). Then the clip was withdrawn into the channel to avoid trauma when insertion of the endoscope (b).
The tunneling method (Figure 3): Mucosal incision of the anal and oral sides of the lesion were performed by a dual knife. After that, a submucosal tunnel was made from the oral side to the anal side by the dual knife. Then, the mucosa on the right side and left side was incised by an IT2 knife to complete a circumferential incision. Finally, the remaining submucosal fibers were dissected and to achieve en bloc dissection.
Figure 3.
Esophageal ESD using the tunneling method on live pigs.
(a) The operative process was as follows: (i) & (ii): Mucosal incision of the anal and oral part was performed. (iii): A submucosal tunnel was made from the oral side to the anal side by submucosal tunneling dissection. (iv) & (v): The bilateral mucosa was incised and the remaining submucosal fibers were dissected.
(b) Mucosal incision of the anal side was performed by a dual knife.
(c) Mucosal incision of the oral side was performed by a dual knife.
(d) A submucosal tunnel was made from the oral side to the anal side by a dual knife.
(e) The mucosa on the left side was incised by an IT2 knife.
(f) The mucosa on the right side was incised by an IT2 knife.
ESD, endoscopic submucosal dissection.
The conventional method (Figure 4): The process was begun by making a semicircular mucosal incision on the right side by a dual knife; followed by submucosal dissection of the right side by the dual knife; then, the same process was repeated on the left sides.
Figure 4.
Esophageal ESD using the conventional method on live pigs.
(a) The operative process was as follows: (i): A semicircular mucosal incision was made on the right side. (ii): Submucosal dissection of the right side was performed. (iii): Semicircular mucosal incision of the left side was performed. (iv): Submucosal dissection of the total submucosa was performed.
(b) A semicircular mucosal incision was made on the right side by a dual knife.
(c) & (d) Submucosal dissection of the right side was performed by a dual knife.
(e) Semicircular mucosal incision of the left side was performed by a dual knife.
(f) Submucosal dissection of the total submucosa was performed by a dual knife.
ESD, endoscopic submucosal dissection.
Outcomes
The primary outcomes were procedural time and speed. The procedural time was defined as the time elapsed from the beginning of mucosal marking to the completion of specimen resection (including the time for treatment of the adverse events during ESD such as hemostasis). The procedural speed (mm2/min) was calculated as the square of the resected specimens (length × width) divided by the procedural time.
The secondary outcomes were en bloc resection rates, adverse events, and self-completion rates. En bloc resection was defined as the removal of the entire simulated lesion in one piece with all marking spots remaining on the resected specimens. Adverse events included perforation and muscle layer injury during the procedures. Muscle layer injury was defined as injury to the muscularis propria but without obvious perforation. Procedures completed by the trainees themselves without taken over by the senior endoscopist were considered as ‘self-completed’.
The outcomes of all 60 procedures performed by the trainees were recorded, and learning curves were obtained according to these results. Moreover, ESDs performed by the senior endoscopists as demonstrations on pigs were also evaluated (3 procedures in each group), which were used to be ‘hypothetical endpoints’ on the learning curves.
Statistical analysis
An analysis of variance was used to compare the procedural time and speed and the size of specimens among different groups. A Chi-squared test was used to compare en bloc resection rates, adverse events, and self-completion rates. SPSS13.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. A statistically significant difference was established when p < 0.05.
Sample size calculation: ESDs done by the first three trainee endoscopists, who were randomized to one per group and each performed four ESDs, were considered as a pilot study. In the pilot study, the procedure time of the clip-with-line method, the tunneling method, and the conventional method was 48.3 ± 5.7, 72.8 ± 4.1, and 75.5 ± 7.0 min, respectively. The adverse event rates were 25% (1/4), 75% (3/4), and 100% (4/4), respectively. Then the sample size was calculated according to the results of pilot study. With a 95% confidence interval, 80% power, and equal sample number in each arm, the minimal number of procedures that could demonstrate that the clip-with-line method is superior to the other two methods in procedure time was 1 in each arm, while the minimal number to demonstrate the superiority in avoiding adverse events was 19 in each arm.
Results
In each group, 20 esophageal ESDs were performed by the trainees. As shown in Table 1, the procedural time in the clip-with-line group was significantly shorter than in the tunneling and the conventional method group (47.4 ± 9.0 min versus 67.0 ± 15.1 min and 67.0 ± 11.8 min, p = 0.000). Though equally sized simulated lesions were marked in each group before dissection, the size of actually dissected specimens were different among each group (p = 0.002). However, the procedural speed resulted in a similar pattern to procedural time: the clip-with-line method had the fastest speed (16.2 ± 4.3 mm2/min), then the tunneling method and then the conventional method, though the latter two had no statistically significant difference (10.5 ± 2.4 and 9.6 ± 2.0 mm2/min, respectively). The en bloc resection rates and self-completion rates were similar among the three groups. The conventional method had a significantly higher rate of perforation than the clip-with-line method and the tunneling method (40% versus 5% and 20%, p = 0.026). The clip-with-line method had the lowest rate of muscle layer injury in the three methods.
Table 1.
Outcomes of ESD performed by trainees in the clip-with-line method group, the tunneling method group, and the conventional method group.
| Clip-with-line | Tunneling | Conventional | p value | |
|---|---|---|---|---|
| Procedures (n) | 20 | 20 | 20 | |
| Procedural time (min)* | 47.4 ± 9.0 | 67.0 ± 15.1 | 67.0 ± 11.8 | 0.000 |
| Size of specimen (mm2)* | 740.7 ± 142.7 | 675.0 ± 72.3 | 622.2 ± 76.4 | 0.002 |
| Procedural speed (mm2/min)* | 16.2 ± 4.3 | 10.5 ± 2.4 | 9.6 ± 2.0 | 0.000 |
| En bloc resection rate | 90.0% (18/20) | 95.0% (19/20) | 85.0% (17/20) | 0.574 |
| Adverse events | ||||
| Perforation | 5.0% (1/20) | 20.0% (4/20) | 40.0% (8/20) | 0.026 |
| Muscle layer injury | 45.0% (9/20) | 75.0% (15/20) | 95.0% (19/20) | 0.002 |
| Self-completion rate | 80.0% (16/20) | 75.0% (15/20) | 75.0% (15/20) | 0.911 |
Data are shown as mean ± SD.
ESD, endoscopic submucosal dissection; SD, standard deviation.
Learning curves of these three methods were drawn according to the outcomes of procedural time and speed (Figure 5). The curves demonstrated that the clip-with-line method achieved higher procedural speed and consumed less procedural time than the other two methods after only 2–3 sessions of ESD training. After four sessions of ESD, the procedural speed of a trainee using the clip-with-line method was almost same to that of the senior endoscopists using the other two methods. The learning curves of tunneling method and conventional method almost coincided.
Figure 5.
Learning curves of esophageal ESD on live pigs. Each trainee performed four ESDs. ESDs performed by the senior endoscopists were used to be the hypothetical endpoints. Data were shown as mean ± SD. *p < 0.05.
ESD, endoscopic submucosal dissection; SD, standard deviation.
Discussion
To our knowledge, this is the first report to compare three different methods performing esophageal ESD on animal models. Our results demonstrated that the clip-with-line method not only significantly shortened the procedural time compared with the tunneling method and the conventional method, but was also associated with lower rates of perforation and muscle layer injury. Learning curves analysis showed the clip-with-line method was easier for the trainees to master than the other two methods. Based on the above results, this clip-with-line method is the preferred method for endoscopists in their learning periods of performing esophageal ESD.
Adequate tissue tension and a clear view of the tissue to be dissected are important for effective and safe ESD. The gravity of the lesion itself and the fluid injected to the submucosal layer are commonly used for improving tissue tension and increasing the visibility of operative field in the conventional method. Recently, various traction methods have been developed to make ESD safer and faster, such as the snare method [Jin et al. 2015], the magnetic anchor method [Gotoda et al. 2009], and the double-scope method [Ahn et al. 2011]. The clip-with-line method is a simple, easy to use, and inexpensive traction method. The feasibility and advantages of the clip-with-line method in gastric [Suzuki et al. 2016; Jeon et al. 2009] and esophageal ESD [Ota et al. 2012; Koike et al. 2015] have been demonstrated by previous clinical case-control studies. This method is especially suitable for esophageal ESD, as the small clip and the thin line do not interfere with endoscopic movement in the narrow space of the esophageal lumen compared with other traction devices. Additionally, though the direction of the traction is only on the oral side, the procedure in retroflex view is extremely rare in the esophageal lumen, making the anal side traction unnecessary. Hence, we adopted this traction method in the current esophageal ESD training course. Moreover, an IT2 knife was used in circumferential incision and submucosal dissection, as the insulation tip of the IT2 knife was safer for beginners to use, and the dissection speed was faster compared with a dual knife.
The tunneling method is another technique that was reported to facilitate esophageal ESD [Linghu et al. 2013; Pioche et al. 2013]. As the lateral edges of mucosa remains intact during submucosal tunnel dissection, the diffusion of injected fluids into both sides of the lesion is prevented which results in longer cushion effect and thereby increases the space between the submucosal / muscular layers in the tunnel and makes it easier to visualize the dissection zone. However, in the current study, the tunneling method showed no significant difference in procedural time compared with the conventional method, although lower rates of perforation and muscle layer injury were observed. As reported in the previous studies [Linghu et al. 2013; Pioche et al. 2013], the tunneling method is more suitable for larger esophageal lesions (i.e. lesions occupied more than half circumference of the esophageal lumen). In our study, however, the simulated lesion was limited as a quarter of the esophageal circumference. Therefore, the advantages of the tunneling method could not be displayed enough in those smaller lesions in our study.
There are very few reports about ESD training courses on animal models. The use of animal models to facilitate the early training of ESD is important to minimize the risk of higher complications at the beginning of the learning curve in humans, this is especially important for countries or regions where the prevalence of upper gastrointestinal neoplasms is low and the opportunities to perform ESD are few [Gotoda et al. 2014]. A European study about hands-on training of gastric ESD with live pigs reported a perforation rate of 22% [Berr et al. 2011]. In our study, the rate of perforation ranged from 5% for the clip-with-line method to 40% for the conventional method. Our results demonstrated that using of animal models allowed endoscopists to ascend the learning curve in a relatively short time with the clip-with-line method. However, the simulated lesions on pigs would be different from humans, such as less frequent bleeding and less fibrosis [Parra-Blanco et al. 2010]. Therefore, the feasibility of the clip-with-line method as the first choice of esophageal ESD for the beginners still need to be confirmed in further clinical studies. In addition, as the sample size was small, we could not analyze the learning curve about complication rates in the current study. The effect of training programs on avoiding complications of ESD still needs further study.
Our study has several limitations. First, the IT2 knife used in the clip-with-line group is faster for submucosal dissection than the dual knife used in the other groups [Oyama, 2014], which might shorten the procedure time independently and lead to a bias of the result. Additionally, the IT2 knife used in our study was reportedly not suitable for esophageal ESD as it is associated with a higher perforation rate [Oyama, 2014]. Although the perforation rate was low for the clip-with-line method, a high rate of muscle layer injury was observed in the current study, which could be partially explained by the use of IT2 knife. Recently, the ITknife nano (KD-612, Olympus) has been developed for colonic and esophageal ESD. Unfortunately, during the study period, the ITknife nano was not available in the Olympus China Medical Training and Education Center where the study was carried out. Second, other adverse events, in addition to perforation and muscle injury, such as bleeding and mediastinal emphysema, was not analyzed. Nevertheless, post-procedural bleeding and mediastinal emphysema are difficult to assess in an animal model. Thirdly, although we tried to create a standardized simulated lesion in each procedure, no two lesions were identical. Individual pigs have their own character such as bleeding easily, thick mucosa and different anatomy. Even two adjacent pieces of mucosa in the same pig would be different. These different characters may have potential impact of the result of this study.
In conclusion, the efficacy and safety of the clip-with-line method in esophageal ESD for trainees was demonstrated in this study on live pigs with standardized setup and expert supervision. Our results revealed that the method is the preferred method for endoscopists in their learning periods of performing esophageal ESD at the present time.
Acknowledgments
Sheng JQ, Fu KI, and Jin P designed the study. Jin P, Fu KI, Yu Y, He YQ, Wei Z, and Wang X organized the project and performed the study. Jin P and Yu Y analyzed the data. Jin P and Fu KI drafted the article, and Cai Q made critical revision for the article. Sheng JQ made the final approval of the article.
Footnotes
Funding: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement: The Olympus China Medical Training and Education Center (Shijingshan, Beijing, China) is a commercial organization which belongs to the Olympus company. The manufacturers were not involved in either the design of the study or the analysis of the data.
Contributor Information
Peng Jin, Department of Gastroenterology, PLA Army General Hospital, Beijing 100700, China.
Kuang-I Fu, Department of Gastroenterology, PLA Army General Hospital, Beijing 100700, China.
Yang Yu, Department of Gastroenterology, PLA Army General Hospital, Beijing 100700, China.
Yu-Qi He, Department of Gastroenterology, PLA Army General Hospital, Beijing 100700, China.
Zhi Wei, Department of Gastroenterology, Jinan Military General Hospital, Jinan, China.
Xin Wang, Department of Gastroenterology, PLA Army General Hospital, Beijing 100700, China.
Qiang Cai, Division of Digestive Diseases, Emory University School of Medicine, Atlanta, GA, USA.
Jian-qiu Sheng, Department of Gastroenterology, PLA Army General Hospital, Nanmencang 5#, Dongcheng District, Beijing 100700, China.
References
- Ahn J., Choi K., Choi J., Kim M., Lee J., Choi K., et al. (2011) Transnasal endoscope-assisted endoscopic submucosal dissection for gastric adenoma and early gastric cancer in the pyloric area: a case series. Endoscopy 43: 233–235. [DOI] [PubMed] [Google Scholar]
- Berr F., Ponchon T., Neureiter D., Kiesslich T., Haringsma J., Kaehler G., et al. (2011) Experimental endoscopic submucosal dissection training in a porcine model: learning experience of skilled Western endoscopists. Dig Endosc 23: 281–289. [DOI] [PubMed] [Google Scholar]
- Cao Y., Liao C., Tan A., Gao Y., Mo Z., Gao F. (2009) Meta-analysis of endoscopic submucosal dissection versus endoscopic mucosal resection for tumors of the gastrointestinal tract. Endoscopy 41: 751–757. [DOI] [PubMed] [Google Scholar]
- Fujishiro M., Kodashima S., Godo O., Ono S., Niimi K., Yamamichi N., et al. (2009) Endoscopic submucosal dissection for esophageal squamous cell neoplasams. Dig Endosc 21: 109–115. [DOI] [PubMed] [Google Scholar]
- Gotoda T., Ho K., Soetikno R., Kaltenbach T., Draganov P. (2014) Gastric ESD: current status and future directions of devices and training. Gastrointest Endosc Clin N Am 24: 213–233. [DOI] [PubMed] [Google Scholar]
- Gotoda T., Oda I., Tamakawa K., Ueda H., Kobayashi T., Kakizoe T. (2009) Prospective clinical trial of magnetic-anchor-guided endoscopic submucosal dissection for large early gastric cancer. Gastrointest Endosc 69: 10–15. [DOI] [PubMed] [Google Scholar]
- Jeon W., You I., Chae H., Park S., Youn S. (2009) A new technique for gastric endoscopic submucosal dissection: peroral traction-assisted endoscopic submucosal dissection. Gastrointest Endosc 69: 29–33. [DOI] [PubMed] [Google Scholar]
- Jin P., Yu Y., Fu K., Yu D., Li A., Sheng J. (2015) A new traction method with use of the snare as a “second hand” during endoscopic submucosal dissection. Endoscopy 47(Suppl. 1): E286–E287. [DOI] [PubMed] [Google Scholar]
- Koike Y., Hirasawa D., Fujita N., Maeda Y., Ohira T., Harada Y., et al. (2015) Usefulness of the thread-traction method in esophageal endoscopic submucosal dissection: randomized controlled trial. Dig Endosc 27: 303–309. [DOI] [PubMed] [Google Scholar]
- Linghu E., Feng X., Wang X., Meng J., Du H., Wang H. (2013) Endoscopic submucosal tunnel dissection for large esophageal neoplastic lesions. Endoscopy 45: 60–62. [DOI] [PubMed] [Google Scholar]
- Maeda Y., Hirasawa D., Fujita N., Suzuki T., Obana T., Sugawara T., et al. (2011) Mediastinal emphysema after esophageal endoscopic submucosal dissection: its prevalence and clinical significance. Dig Endosc 23: 221–226. [DOI] [PubMed] [Google Scholar]
- Ota M., Nakamura T., Hayashi K, Ohki T., Narumiya K., Sato T., et al. (2012) Usefulness of clip traction in the early phase of esophageal endoscopic submucosal dissection. Dig Endosc 24: 315–318. [DOI] [PubMed] [Google Scholar]
- Oyama T. (2014) Esophageal ESD: technique and prevention of complications. Gastrointest Endosc Clin N Am 24: 201–212. [DOI] [PubMed] [Google Scholar]
- Oyama T., Kikuchi Y., Shimaya S. (2002) Endoscopic mucosal resection using a hooking knife (hooking EMR). Stomach Intest 37: 1155–1161. [Google Scholar]
- Oyama T., Tomori A., Hotta K. (2006) ESD with a hook knife for early esophageal cancer. Stomach Intest 41: 491–497. [Google Scholar]
- Parra-Blanco A., Arnau M., Nicolas-Perez D., Gimeno-Garcia A., Gonzalez N., Diaz-Acosta J., et al. (2010) Endoscopic submucosal dissection training with pig models in a Western country. World J Gastroenterol 16: 2895–2900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pioche M., Mais L., Guillaud O., Hervieu V., Saurin J., Ponchon T., et al. (2013) Endoscopic submucosal tunnel dissection for large esophageal neoplastic lesions. Endoscopy 45: 1032–1034. [DOI] [PubMed] [Google Scholar]
- Rosch T., Sarbia M., Schumacher B., Deinert K., Frimberger E., Toermer T., et al. (2004) Attempted endoscopic en bloc resection of mucosal and submucosal tumors using insulated-tip knives: a pilot series. Endoscopy 36: 788–801. [DOI] [PubMed] [Google Scholar]
- Suzuki S., Gotoda T., Kobayashi Y., Kono S., Iwatsuka K., Yaqi-Kuwata N., et al. (2016) Usefulness of a traction method using dental floss and a hemoclip for gastric endoscopic submucosal dissection: a propensity score matching analysis (with videos). Gastrointest Endosc 83: 337–346. [DOI] [PubMed] [Google Scholar]
- Teoh A., Chiu P., Wong S., Sung J., Lau J., Ng E. (2010) Difficulties and outcomes in starting endoscopic submucosal dissection. Surg Endosc 24: 1049–1054. [DOI] [PubMed] [Google Scholar]