Efficacy of innovative polyglycolic acid sheet device delivery station system: a randomized prospective study

Background Although there have been several reports of treating large post-endoscopic submucosal dissection (ESD) ulcers by covering them with a polyglycolic acid sheet (PGAs), this approach presents problems regarding PGAs delivery. This study assessed the usefulness of a device delivery station system (DDSS) to evaluate the appropriate and rapid PGAs coating method with DDSS. Methods Thirty-nine of 41 patients who were diagnosed with early gastric cancer over 20 mm in diameter and pathologically diagnosed with well-differentiated adenocarcinoma were randomly allocated to the following two groups according to delivery method: the conventional PGAs delivery group (C group) (n = 19) and the new DDSS group (DDSS group) (n = 20). The primary outcome was the coating area per minute in the C group and DDSS group (cm2/min). Results There were significant differences in the coating time (min), with values of 34.1 (15.0–60.7) vs. 16.85 (11.5–27.2) min for the C group and DDSS group, respectively (p = 0.001). There was also a significant difference in coating area per minute, with values of 0.261 (0.02–1.00) and 0.96 (0.173–2.06) cm2/min for the C group and DDSS group, respectively (p = 0.001). There were four cases of post-ESD bleeding (1–7 days after ESD) in the C group compared with 0 in the DDSS group, which represented a significant difference (p = 0.030). Conclusions The DDSS was very useful for rapidly delivering and tightly attaching a PGAs to control post-ESD bleeding. Trial registration University Hospital Medical Network (UMIN) 000026377. Electronic supplementary material The online version of this article (10.1007/s00464-017-6019-6) contains supplementary material, which is available to authorized users.

As such, filling and covering large artificial defects with a PGAs is a more ideal treatment approach; however, it is difficult to deliver and thoroughly cover an ulcer floor with a thin PGAs using only a pan-endoscope, especially under wet conditions. PGAs delivery is one of the main problems of treating artificial ulcers. The conventional PGAs coating method is very time consuming and causes maneuverability difficulties [4][5][6][7][8]. As a 30-to 40-mm square-shaped PGAs is gripped by grasping forceps through the endoscopic channel and delivered from the mouth to the stomach along with the endoscope, the PGAs is exposed to saliva and gastric juices, and it is difficult for the PGAs to maintain its sheet shape under such wet conditions. Once a PGAs becomes curled up in the stomach, it is very difficult if not impossible to uncurl the PGAs into its sheet form using an endoscope [9,10]. This study aimed to evaluate the efficacy of appropriate and rapid PGAs coating method using device delivery station system (DDSS).

Patients and methods
This study was conducted with approval from the ethics committees of the Ehime Rosai and Kagawa University Hospitals (Approval No. 80) and in accordance with the Declaration of Helsinki. The patients provided verbal and written informed consent. This trial was also registered with the University Hospital Medical Information Network (UMIN000026377) following the CONSORT check list.
Forty-one patients who were diagnosed with early gastric cancer over 20 mm in diameter by narrow band imaging magnified endoscopic examination (NBI-ME) and pathologically diagnosed with well-differentiated adenocarcinoma from biopsy specimens were eligible for this prospective randomized study. Two patients refused to participate and were excluded. The remaining 39 patients were randomly allocated using the sealed envelope method into the following two groups according to the delivery method: the conventional PGAs delivery group (n = 19) (C group) and the new DDSS group (n = 20) (DDSS group) (Fig. 1). Patients taking ticlopidine hydrochloride, clopidogrel sulfate or aspirin were switched to cilostazol 3 days before ESD, and they discontinued cilostazol the day of ESD. All antiplatelet drugs were resumed the day after ESD. The three endoscopists who performed ESD were members of the Japan Gastroenterological Endoscopy Society and received a lecture on the DDSS beforehand.
In the C group, a 50-mm square-shaped PGAs was gripped by grasping forceps through an endoscopic channel and delivered from the mouth to the stomach along with the endoscope. The PGAs coating time was defined as the duration from the beginning of endoscope insertion into the mouth to the end of the fibrin glue coating process.
In the DDSS group, a nasal endoscope with an attached DDSS was prepared beforehand following specified procedures. The PGAs coating time was defined the same as it was for the C group. In addition to several parameters measured during ESD, other calculations and analyses were conducted as follows: The short axis (S) and long axis (L) (cm) of the ellipsoid dissected specimen were measured after ESD.
• The ellipsoid coating area was defined as the area calculated by the following formula: the ellipsoid coating area (cm 2 ) = π × L/2 × S/2 (π = 3.14) • The coating area per minute was calculated by the following formula: the coating area per minute (cm 2 /min) = the ellipsoid coating area (cm 2 )/the PGAs coating time (min)

Primary outcome
1. Coating area per minute in the C group and DDSS group, respectively (cm 2 /min).

Sample size calculation
After we conducted a pilot study of five patients in the C group and five patients in the DDSS group (a total of ten patients), we found significant differences between the two groups in the coating area per minute (C) (C C and C DDSS ) (cm 2 /min). Based on these results, the sample size required for an α error of 0.05 and a power of 0.8 with a standard deviation of 0.0351 was calculated to be 20 patients by performing a statistical analysis using Graph-Pad Prism (http://biostat.mc.vanderbilt.edu/wiki/Main/ PowerSampleSize).

Video 1
An endoscopic injection sclerotherapy (EIS) balloon (MD-47411L, 8 mm in diameter, 60 mm in length) (Sumitomo Bakelite Co., Tokyo, Japan) and a nasal endoscope (GIF TYPE XP260NS, Olympus Co., Tokyo, Japan) were prepared. Ring-shaped threads 3-4 mm in diameter of different colors were placed in each corner of the 50-mm PGAs. A 5-cm thread was attached to one side of the square PGAs and was used for housing PGAs in the EIS balloon ( Fig. 2A). The nasal endoscope was inserted into the EIS balloon, and the 5-cm thread attached on PGAs was pulled through the EIS balloon (Fig. 2B). PGAs was stored in the gap between the nasal endoscope and the EIS balloon (Fig. 2C). PGAs was distributed equally around the nasal endoscope ( Fig. 2D). PGAs was equally distributed around the nasal endoscope ( Fig. 3A, B). Approximately 4.5 ml of air was insufflated, which did not cause outer EIS balloon swelling (Fig. 3C). Thus completing the preparation of the nasal endoscope with PGAs delivery system (NE-PGAs) was made (Fig. 3D).

Video 2
Esophagogastroduodenoscopy revealed rather thick, 0-IIa-type, early gastric cancer 45 mm in diameter in the posterior wall of the body of the stomach (Fig. 4A). The post-ESD artificial ulcer became 50 mm in diameter (Fig. 4B). NE-PGAs was inserted from the mouth to the stomach (Fig. 4C). After 4.5 ml of air was deflated from the EIS balloon, the nasal endoscope was pulled out through the EIS balloon (Fig. 4D). By pulling out the nasal endoscope slowly, PGAs was observed (Fig. 4E). By grasping the thread, PGAs could be pulled out from the NE-PGAs (Fig. 4F). PGAs was brought to the post-ESD ulcer ( Fig. 5A). PGAs was unfolded by referencing the four threads attached to the corners (Fig. 5B). All of the threads attached to PGAs were clipped and fixed equally (Fig. 5C). More clips were added just above the ulcer floor ( Fig. 5D). Beriplast P Combi-Set Tissue Adhesive®(CSL Behring K.K., Tokyo, Japan) (combination of fibrin glue and thrombin) was applied equally to the artificial ulcer (Fig. 5E). The tight attachment of PGAs to the post-ESD artificial ulcer floor was completed (Fig. 5F).

Statistical analysis
For comparing the relative frequencies between groups, data were analyzed using Fisher's exact test or the χ 2 test. The Mann-Whitney U test was used to compare continuous variables with a significance level of p < 0.05. Statistical analyses were performed using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, CA, USA).

Results
There were no significant differences in patient age or gender (p = 0.66 and 0.848, respectively) between the C group and DDSS group. Regarding the location of lesions, in the C group (19 lesions), six were upper (U) lesions, seven were middle (M) lesions, and six were lower (L) lesions. In the DDSS group (20 lesions), five were U lesions, eight were M lesions, and seven were L lesions. There was no significant difference in lesion location (p = 0.209). The macroscopic findings of lesions revealed that the 0-IIa, 0-IIc, and 0-IIa + IIc types accounted for 6, 4, and 9 lesions, respectively, in the C group. In the DDSS group, the 0-IIa, 0-IIc, and 0-IIa + IIc types accounted for 7, 5, and 8 lesions, respectively. There was no significant difference in the macroscopic findings (p = 0.479) ( Table 1).
There was also no significant difference in the ellipsoid dissected area (cm 2 ), with areas of 16.9 (3.68-25.3) cm 2 and 15.3 (3.56-27.2) cm 2 for the C group and DDSS group, respectively (p = 0.555) ( Table 2) (Fig. 6A). There was a Fig. 2 PGAs with ring-shaped threads of four different colors. A A PGAs was cut into a 50-mm square sheet. Ring-shaped threads 3-4 mm in diameter of four different colors were placed in the corners of a 50-mm square sheet. A ring-shaped thread was also placed in the center of PGAs. The location of each colored thread is important for recognizing the orientation of PGAs, especially if PGAs becomes wet and curled up in the stomach due to unpredicted fac-tors. For example, the thread on the contralateral side to the black thread was green, and that to the white thread was purple. B A nasal endoscope was inserted into an endoscopic injection sclerotherapy (EIS) balloon, and the 5-cm thread on one side of the square PGAs was pulled through the EIS balloon. C PGAs was stored in the gap between the nasal endoscope and the EIS balloon by pulling the thread. D Then, PGAs was placed equally around the nasal endoscope significant difference in the coating time (min), with values of 34.1 (15.0-60.7) and 16.85 (11.5-27.2) min for the C group and DDSS group, respectively (p = 0.001) (Fig. 6B). In the coating areas per minute (cm 2 /min), there was a significant difference, with values of 0.261 (0.02-1.00) cm 2 / min and 0.96 (0.173-2.06) cm 2 /min (p = 0.001) ( Table 2) (Fig. 6C) for the C group and DDSS group, respectively.
Anticoagulants were used in four and five patients in the C group and DDSS group, respectively (p = 0.777). Antiplatelet drugs were used in three and four patients in the C group and DDSS group, respectively (p = 0.740). There were no significant differences between the groups (Table 1).
Among antiplatelet drugs, aspirin/ticlopidine hydrochloride/clopidogrel sulfate were taken by 2/3/2 patients in the C group and by 3/4/2 patients in the DDSS group, respectively. There was no significant difference in the use of antiplatelet drugs between the groups (p = 0.383) ( Table 1). There were 4 (21% of C group) cases of post-ESD bleeding (1-7 days after ESD) in the C group compared with 0 in the DDSS group, which was significantly different (p = 0.030). There were 0 cases of perforation during ESD in both groups, with no significant difference (p = 0.725) ( Table 2). Instances of post-ESD bleeding were successfully controlled with hemostatic forceps. Total seven patients in C groups and eight patients in DDSS groups allowed us to perform follow up EGD examination. Patients those PGAs remained until postoperational day (POD) 3, 7, 14, 30, and 60 were 4/2/1/0/0 patients in C groups, and 1/2/5/0/0 patients in DDSS groups with significant difference (p = 0.0006). The typical healing course of the PGAS-coated post-ESD ulcer floor in the DDSS group is shown in Fig. 7. Images were captured of the PGAS-coated ulcer floor just after ESD (Fig. 7A) and on POD 3 (Fig. 7B), POD 7 (Fig. 7C), POD 14 (Fig. 7D), POD 30 (Fig. 7E), and POD 60 (Fig. 7F).
Pathological examinations revealed no significant differences in early cancer invasion depth, i.e., mucosal, Fig. 3 PGAs stored equally around the nasal endoscope by air insufflation. A Using tweezers, PGAs was adjusted to be placed equally around the nasal endoscope. B As viewed from in front of the endoscope tip, the PGAS was placed in the gap between the EIS and nasal endoscope. C As EIS balloons swell equally outward and inward, PGAs was sealed between the endoscope and the inner bulge of the swollen balloon, then, insufflating approximately 4.5 ml of air did not cause the outer balloon to swell and PGAs was sealed in the center of an EIS balloon. D Thus completing the preparation of the nasal endoscope with PGAs delivery system (NE-PGAs) was made submucosal or sm massive, with 13, 4, and 2 cases in the C group and 15, 3, and 2 cases in the DDSS group showing mucosal and submucosal invasion depths, respectively (p = 0.593) ( Table 2).

Discussion
PGAs are made of PGA, which is widely used in bioabsorbable sutures in the surgical field. During the healing process of a post-ESD artificial ulcer, a PGAs exerts anti-inflammatory effects and prompts the creation of rich granulation tissue. Simultaneously, the PGAs prompts the migration of epidermal cells over the rich granulation tissue and accelerates the activity of fibroblasts to form collagenous tissue with a scar. After these processes, the PGAs becomes almost completely absorbed within approximately 3 months (15 weeks). Early anti-inflammatory effects and rich granulation tissue formation not only prevent deformation of the post-ESD artificial ulcer by fibroblasts but also play a role in protecting the ulcer floor from exogenous materials and factors [11]. These protective mechanisms seem to be very similar to reactions induced by steroid local injection therapy to prevent stenosis or deformation of the esophagus or stomach [12][13][14]. Although a PGAs has no inherent hemostatic activity, covering the ulcer floor with a PGAs and coating the covering with fibrin glue provide protection, deformity prevention, and hemostatic effects. The tight attachment of a PGAs to the artificial floor is necessary to obtain a sufficient hemostatic effect; insufficient attachment leads to post-ESD bleeding. PGAs is very useful for covering post-ESD ulcer floors to provide protection from exogenous materials and hemostasis with an additional fibrin glue coating [15,16]. air was deflated from the EIS balloon and the nasal endoscope was pulled out through the EIS balloon with PGAs. E By pulling out the nasal endoscope slowly, the white and black threads attached to PGAs corners could be observed. While the EIS balloon and PGAs remained in the stomach, the nasal endoscope was replaced by an oral endoscope to proceed with the treatment. F By grasping the thread attached to the corner of PGAs, PGAs was pulled out from NE-PGAs  On the other hand, there are two major problems in applying a PGAs to a post-ESD ulcer floor, as follows: ① rapidly delivering the thin PGAs without exposing it to saliva or gastric juice [10]; and ② tightly attaching the PGAs to the ulcer floor using only a flexible endoscope such that a sufficient hemostatic effect is obtained [9]. For ①, we have developed a novel endoscopic device delivery concept and produced a prototype of the innovative device, i.e., the E-DDSS (Fig. 8A). The E-DDSS can deliver various endoscopic devices that can be placed within the digestive tract via a detachable functionality under completely sealed conditions (Fig. 8B). E-DDSS has two chambers in which PGAs were stored, and it was delivered by splitting one side of the device by unlocked system (Fig. 8C).  There were two types of E-DDSS prototypes (3 or 5 cm in length) (Fig. 8D). From the E-DDSS attached to the endoscope, it is possible to pull out items loaded in the circumferential cavities, such as hemostatic gauze or a PGAs (Fig. 8E). The E-DDSS can be attached to an oral endoscope or a nasal endoscope (Fig. 8F). The E-DDSS is now being tested in several in vivo experiments. As a DDSS that can deliver a PGAs is included within one of the concepts of the E-DDSS, in this study, a DDSS that used an existing EIS balloon was used to deliver a PGAs, and this approach yielded better results than the conventional delivery method. However, using the DDSS with a nasal endoscope for PGAs delivery (NE-PGAs) required the immediate replacement of the nasal endoscope by an oral endoscope, which slightly increased the procedure duration [10]. For ②, tight PGAs attachment was achieved by using a PGAs with four ring-shaped colored threads placed in the four corners of the sheet to more easily recognize the orientation of the PGAs and facilitate sheet fixation with hemoclips. Post-ESD bleeding is one of the crucial events especially for patients who take antiplatelet agents. There were some reports that post-bleeding events occurred within 7-14 days after ESD. Follow-up EGDs until POD 60 revealed PGAs remained on POD 7-14 in almost of all patients without post-ESD bleeding events. Using a PGAs with four ring-shaped colored threads also enabled the effective, tight attachment of the PGAS such that post-ESD bleeding was decreased among patients using antithrombotic agents [17][18][19]. In our study, with limitation of small number, although PGAs prevented post-ESD bleeding, it could not prevent the deformity of post-ESD ulcer. Although there were several reports that PGAs prevented stricture of esophagus after ESD [3,11], we could not confirm anti-deformity effect of PGAs. Considering the healing process, PGAs prompts the migration of epidermal cells and accelerates the activity of fibroblasts to form collagenous tissue with a scar, and these effects might be able to prevent post-ESD bleeding, but not deformity of stomach. Further multi-center studies are In conclusion, this DDSS was very useful for rapidly delivering and tightly attaching a PGAs to control post-ESD bleeding.

Fig. 8
Concept and prototype of the endoscopic device delivery station system (E-DDSS). A The E-DDSS can deliver various endoscopic devices and place them within the digestive tract via a detachable functionality while maintaining completely sealed conditions. B E-DDSS can deliver various endoscopic devices and detained within the digestive tract by detachable function under completely sealed condition. C E-DDSS has two chambers in which PGAs were stored, and it was delivered by splitting one side of the device by unlocked system. D Two E-DDSS prototypes (3 or 5 cm in length). E From independent E-DDSS attached to the endoscope, it is possible to pull out loaded items, such as hemostatic gauze or a PGAs. F The E-DDSS can be attached to an oral endoscope or a nasal endoscope