This prospective, single-center randomized controlled trial was designed to compare the procedure-related outcomes between GTS-ESD and C-ESD when treating superficial gastric neoplasms. The trial was approved by the Ethics Committee of Zhejiang Provincial People’s Hospital, Chun’an branch (Ethics Approval Number: 2022-03-10-05) and registered with the National Medical Research Registration Information System (www.yxyj.org.cn; identification No.:MR-33-24-020319), conducted following the Declaration of Helsinki. All participants provided written informed consent for this study before enrollment.

Consecutive patients who underwent ESD for superficial gastric neoplasms between November 2022 and May 2024 were enrolled in this study. The gastric ESD procedures were performed by a single endoscopist whose experience amounted to 100 procedures per year. The eligibility criteria were as follows: (1) age 20 to 90 years old; (2) normal anatomy of the upper digestive tract; (3) endoscopic diagnosis of precancerous lesion, gastric adenomas with high risk of canceration; clinical diagnosis of intramucosal cancer (cT1a), according to the Japanese guidelines for the diagnosis and treatment of gastric cancer [8]; (4) single lesion and diameter ≥ 1 cm; (5) provided informed consent. The exclusion criteria were: (1) lesion near a previous endoscopic resection scar; (2) history of surgery of the upper digestive tract; (3) patients with nasopharyngeal tumors, stenosis, surgical history, and other gastric tube insertion difficulties; (4) severe heart, liver and kidney dysfunction; (5) severe mental disorder; (6) pregnancy or breastfeeding; (7) anesthesia intolerance; (8) anticoagulants or antiplatelet drugs were not discontinued.

Patients were randomly assigned to the conventional ESD (C-ESD) or per-nasal “GTS partner” traction-assisted ESD (GTS-ESD) group in a 1:1 ratio with the computer-based random number table method. The random sequence was stored in assignment tables, following allocation concealment before grouping. Patients were blinded to treatment group assignment, while the endoscopist and investigator were not blinded.

The sample size was calculated based on ESD procedure time. In our preliminary pilot study, the mean procedure time was 42.7 min in the GTS-ESD group and 70.3 min in the C-ESD group, with a common standard deviation of 26.2 min for both groups. In order to ensure a power of 90% with a two-sided error rate of 5%, 40 participants needed to be enrolled. Considering a dropout rate of approximately 20% and other unpredictable factors, the final sample was to include 80 participants.

All patients completed ESD preoperative evaluation, and all patients were fasted for at least 6 h, along with oral administration of defoaming agent (dimethicone) and mucosal dissolving agent (pronase Granules) 20 min prior to ESD. The following operating equipments and materials were used in the study: endoscopy (GIF-Q260J, Olympus) Dual knife (MK-T-2-195, Micro-Tech), injection needles (ATE-ZSZ-C-23 × 2000 × 25 × 4, Atete), hot biopsy forceps (HBF-23-2000, Micro-Tech), clip(ROCC-D-26-195-C, Micro-Tech), snare (SD-T-2423–35, Kangjin), gastric tube (5.3 mm (Fr16), Jiangsu), high-frequency generator (ERBEVIO 200D), carbon dioxide insufflation (Olympus), and submucosal injection was prepared by mixing normal saline 100 ml, methylene blue 0.1 ml, sodium hyaluronate 8 ml, and epinephrine 1 ml.

All patients underwent procedures under monitored anesthesia care with endotracheal intubation. The equipments, materials, and other conditions used in the procedure were identical in both groups except for the traction method in the dissection phase (Fig. 1).

Per-nasal “GTS partner” assisted traction technique. A The head end of the snare is inserted into the side hole of the gastric tube. B The snare is guided by a gastric tube and delivered to the vicinity of the lesion under endoscopic surveillance. C The snare grip is fixed with adhesive tape. D-F A complete circumferential incision of the lesion is completed. Then, one hemoclip is placed on the appropriate lesion margin while another hemoclip is used to orientate the snare, grab the first hemoclip, and expose the submucosal layer by pushing the snare. G–I Lesions from different parts of the stomach, snare tightening lesions directly, exposing the submucosa and maintaining tissue tension by pulling and pushing the snare

Mucosal marking was performed approximately 5 mm outside the margin of the lesion with a Dual knife. Next, submucosal injection was injected into the submucosal layer until the lesions were fully elevated, and the mucosal and submucosal layer surrounding the lesion was dissected along the marked points using a Dual knife. The traction method, underwater technique, or any other special technique [9] was not used in the C-ESD procedure group. The vision of operation was maintained only through the transparent cap during dissection. For the Per-nasal “GTS partner” assisted traction group (Fig. 3B), after a circumferential mucosal incision of the lesion, the assistant inserted the snare into the side hole of the gastric tube and near the lesion through one side nasal cavity (Figs. 2A, B and 3A). The snare was withdrawn from the side hole of the gastric tube, the gastric tube was pulled out, and the snare was opened, after which a clip was inserted into the gastroscopic biopsy hole. The clip grabbed the snare to trap the mucosal edge of the rimb of lesion, and the snare was tightened (Fig. 2G, Video 1). The endoscopist pulled the external snare cannula, which could be pushed or pulled to change the traction direction, and the surgical field was exposed after traction; the tissue tension was maintained during the dissection procedure (Fig. 2F–I). If the first traction did not provide satisfactory visualization of the submucosal layer, the snare was released to select another suitable site for traction. After dissection, hemostatic forceps were used to coagulate the exposed small blood vessels in the wound to prevent postoperative bleeding, and the clip was administered at the suspected injury or defect of the muscle layer. Finally, the gastric tube was indwelled for decompression and drainage. Any traction-related tear was considered traction-related damage to the specimen (Fig. 2).

Schematic diagram of per-nasal “GTS partner” assisted traction technique. A The snare is guided by a gastric tube and delivered to the vicinity of the lesion under endoscopic surveillance. B The snare and the endoscope do not enter the gastric cavity through the same channel, which facilitates the independent movement of the endoscope and the snare and improves the accuracy of endoscopic submucosal dissection

According to the lesion location, size, and intraoperative conditions, the patients fasted for 24–48 h after ESD. Proton pump inhibitors, gastric mucosal protective agents, and nutritional support were routinely used, and prophylactic antibiotics were given if there was intraoperative injury of muscularis propria. The patients were monitored for bleeding, perforation, fever, abdominal pain, and other adverse events through observation of clinical symptoms and laboratory examinations. For patients with early gastric cancer, it was determined whether they needed to undergo additional surgery, review gastroscopy, and abdominal CT according to the Japanese “eCura system” [10].

The general data of ESD perioperative period were observed and compared between the two groups, including age, sex, lesion location, lesion position, lesion diameter, macroscopic type, status of Helicobacter pylori infection, procedure time, resection speed, en bloc resection rate, R0 resection rate, specimen size, delayed bleeding, perforation, other adverse events, histological evaluation, and factors associated with per-nasal “GTS partner” assisted traction (snare disengagement, snare damage to specimen). The main endpoints were ESD procedure time and dissection speed. Secondary endpoints were en bloc resection rate, R0 resection rate, and procedure adverse events rate. Subgroup analysis of gastric ESD procedure time and dissection speed was performed according to lesion location and lesion diameter.

The lesion location (i.e., the upper, middle, or lower) and position (i.e., greater curvature, lesser curvature, anterior wall, posterior wall) were defined according to the Japanese classification of gastric cancer [11]. The procedure time (minutes) was defined as the time from the first injection to the end of tumor dissection. The specimen area was calculated using the ellipse formula: specimen area [mm2] = (transverse diameter/2) × (longitudinal diameter/2) × 3.14 (π). The dissection speed was defined as the specimen area divided by the procedure time. En bloc resection was defined as endoscopic resection of an entire lesion in a single specimen without the use of snare. R0 resection was defined as endoscopic resection of an entire lesion in a single specimen with no histopathological evidence of tumor residue at the resection margins. Cases in which the lesion’s margins were unclear because of electrosurgical or mechanical damage were classified as non-R0 resections. Perforation was diagnosed if intra-abdominal organs or fat tissues were observed during ESD procedure or free air was found on post-procedure radiographs or computed tomography images. Delayed bleeding was defined as clinical symptoms and laboratory changes indicative of bleeding after ESD [12, 13].

SPSS 26.0 software was used to perform all statistical analyses. Continuous data with normal distribution are presented as means and standard deviations (SD) and analyzed using Welch’s t-test. Variables with non-normal distribution are presented as medians and interquartile ranges (IQR) and analyzed using the Wilcoxon rank sum test. Categorical data are described using percentages and analyzed using the chi-squared or Fischer’s exact test. P < 0.05 was considered statistically significant.