Endoscopic resection has been increasing utilized as the treatment for small size gastrointestinal stromal tumors (GIST), of which the best resection method has not been identified. We aim to compare the outcomes of endoscopic full thickness resection (EFTR) versus submucosal tunnelling endoscopic resection (STER) for clinical small gastric GIST. We hypothesize that EFTR could achieve better complete margin negative resection than STER without increase in adverse event.

This is an international multi-center double blinded randomized controlled trial involving four high volume centers from Hong Kong, mainland China, India and Japan. Adult patients with clinical 1.0-3.Scm gastric GIST undergoing endoscopic resection would be recruited.

Patients would be randomized to undergo EFTR (intervention) or STER (Control) by expert endoscopists under general anaesthesia according to well published methods.

Gastrointestinal stromal tumor (GIST) is the most common mesenchymal tumors in the GI tract, often located in the stomach. Based on the latest World Health Organization (WHO) classification, all GISTs are now considered as malignant tumors. Large size overtly aggressive GISTs are relatively rare, occurring only in up to 8 per million population. However, smaller sizes GIST in the stomach are relatively common, and was found in up to 20% of patients based on autopsy series.

Conventionally, localized GISTs are treated by surgical resection. Several guidelines recommended resection of all histologically confirmed GIST, while some suggested surveillance if the lesion is small <2cm in size. The principle of surgery for GIST is for en-bloc margin negative complete resection, while lymph node dissection is not required. As such, laparoscopic resection of gastric GIST has been advocated when technically feasible, demonstrating short term benefits in recovery than open surgery, with similar oncological outcomes.

With the technological advances of endoscopic surgery including endoscopic submucosal dissection (ESD) for early epithelial cancers and per-oral endoscopic myotomy (POEM), there was a rapid expansion in the indication of endoscopic surgery, in particular resection of subepithelial tumors (SET) in the gastrointestinal tract, in which a significant proportion are GISTs. Systematic review revealed a shorter procedure time and improved short-term recovery by endoscopic resection versus laparoscopic resection, without significant difference in complication and survival. With the favourable outcomes consistently reported in the literature regarding endoscopic resection of upper gastrointestinal GISTs, several endoscopy and oncology society guidelines are now recommending endoscopic resection as an option for smaller size GISTs in institutions with expertise on therapeutic endoscopy.

Submucosal tunneling endoscopic resection (STER) was first reported by Xu, et al in 2012. The concept of the procedure is to create a submucosal tunnel away from the tumor that arose from the musclaris propria layer while protecting the mucosa directly overlying the lesion, so that only mucosal closure of the tunnel entrance would be required after resection. The technique was first used on esophageal SET, where majority of them are benign leiomyoma. STER was then subsequently applied to other upper gastrointestinal tract lesions including the stomach. The merit of the technique mainly lies in the simplicity of closure of the mucosal incision, which only requires simple through-the-scope (TTS) clips. A schematic diagram of the STER procedure is shown in Figure 1.

Development of various techniques that allowed secure endoscopic water-tight closure of full thickness wall defect has led to increasing application of endoscopic full thickness resection (EFTR). As opposed to the STER procedure, the tumor would be directly resected without creation of a submucosal tunnel. This would create a full thickness defect that required complete closure to avoid gastrointestinal leakage and peritonitis. Various methods have been reported for closure, ranging from simple TTS clip closure, over-the-scope clip closure, clip endo-loop purse string technique, re-openable clip over-the-line method (ROLM), endoscopic suturing etc. With appropriate selection of closure method based on the morphology of the defect, secure closure could be achieved with minimal post-procedural morbidity.

American Society of Gastrointestinal Endoscopy (ASGE) has recently published a guideline on endoscopic full thickness resection, where STER procedure would be classified as exposed tunneled type EFTR. The EFTR procedure mentioned in previous paragraph would be classified as exposed non-tunneled type EFTR. Due to the complexity of the nomenclature, EFTR and STER will be used in the subsequent text for easier understanding of the technique described.

Both EFTR and STER has been increasingly utilized in resecting gastric subepithelial tumors including GISTs. In a recent systematic review of 952 gastric EFTR procedures including 523 GISTs, en-bloc margin negative resection was achieved in 99.3%, with surgical conversion rate of 0.09%. Pooled estimate of major adverse event was only 0.29%. On the other hand, systematic review of 2941 STER procedure reported margin negative resection rate of 92.4% with major adverse event of 1.2%. Of note, when only gastric lesion or lesion arising from muscularis propria layer were considered, the margin negative resection rate dropped to 90.6% and 88.3% respectively. While both procedures remained safe and feasible, margin negative resection appeared to be better achieved with EFTR. In the aforementioned studies, recurrence was observed on 0% and 2.3% of patients after EFTR and STER respectively.

The investigators have recently reported a retrospective analysis comparing EFTR and STER for gastric GISTs. In line with the current literature, complete margin negative resection was achieved in a significantly higher proportion with EFTR than STER (100% versus 80%, p=0.029), while no difference was found in the incidence of post-procedural adverse event. It is believed that EFTR is superior to STER in obtaining clear surgical margin, as dissection within the submucosal tunnel is challenging in achieving a wide margin without breaching tumor capsule, especially when tumor size is larger than 2cm. The concern for inadequate defect closure has also recently been overcome by numerous new developments of full thickness closure methods as described above. In the study, local recurrence was observed in 1 patient after STER, while no recurrence was found in the EFTR group. The event rate of recurrence was both low for both STER and EFTR, thus statistically significant difference could not be detected without a huge sample size. Nonetheless, it is however anticipated that with better margin negative resection, EFTR could achieve a lower recurrence rate for malignant GIST than STER, and margin negative complete resection should be a reasonable surrogate outcome for oncological clearance.

To date there has not been any prospective comparative study comparing EFTR and STER for small size gastric GISTs. The investigators have therefore designed the current international prospective randomized controlled trial aiming to demonstrate the superiority of EFTR in achieving better margin negative resection.

Consecutive adult patients age >=18 and <=75 with a clinical diagnosis of gastric GIST who opted for endoscopic resection would be screened for eligibility.

Inclusion Criteria:

• Presence of gastric subepithelial tumor on diagnostic upper endoscopy, and
• Diagnostic EUS and CT scan with intravenous contrast suspicious of GIST arising from muscularis propria layer, size with maximum diameter >=1.0cm and <= 3.5cm, and
• Absence of high risk features, including irregular margins, invasion to surrounding organs, lesion hypervascularity, and
• Endoscopic morphology and location deemed feasible with both EFTR and STER by an expert endoscopist, or
• Histological confirmation of GIST through EUS guided fine needle biopsy (Optional, based on recommendation from guidelines)

Exclusion Criteria:

• Patients with tumors deemed not suitable for endoscopic resection (Either EFTR or STER), due to unfavourable location, high risk morphology, or any other reasons.
• Patients with multiple tumors.
• Patients unable or unwilling to provide consent.
• Previous esophageal or gastric surgery.
• Patients with significant cardiorespiratory comorbidities which may limit their ability to undertake general anesthesia for the procedure, including ASA grade III or above.
• Pregnant women or those planning pregnancy or breastfeeding women.
• Uncorrectable coagulopathy defined by international normalized ratio (INR) > 1.5 or platelet count < 50000/µl.
• Patients on double anti-platelet agents or anti-coagulation (Warfarin, heparin or other direct oral anticoagulants)