The clinical data of STMF patients who accepted the new FRW technique at the Fourth Hospital of China Medical University from May 2020 to March 2021 were retrospectively analyzed. This study has been approved by our hospital’s Ethics Committee (Ethics Approval No.: EC-2021-HS-001). All patients have been informed about the FRW technique and given their written consent before the procedure. This study was a retrospective study, not a clinical trial, so we did not register it on clinicaltrials.gov or other similar websites. All authors had access to information that could identify all the patients during or after data collection.

Case selection requirements: (i) gastroscopy revealing elevated mucosa at the fundus of the stomach; (ii) endoscopic ultrasonography revealing STMF; (iii) no history of gastric cardiac or gastric surgery; (iv) tolerance to general anaesthesia and tracheal intubation; (v) lesions diameter of ≤5 cm (21); (vi) patients who were informed with the new endoscopic technique and signed the consent form.

Exclusion Criteria (contraindications for endoscopic resection of gastrointestinal submucosal tumors (21)) were: (i) when an enlarged lymph node or distant metastasis lesion was identified; (ii) patients with a poor general health condition who could not tolerate endoscopic surgery; (iii) patients with bleeding and/or ulceration on the surface of the lesion; (iv) patients whose pathology results showed they were non-GIST patients.

The FRW technique enables the gastroscope to pass through the gastric cardia without entering the gastric antrum while the endoscopy progresses continuously. The gastroscope body makes a forward return along the wall of the posterior wall of the upper gastric body and the wall of greater curvature to access the fornix of the stomach. When the gastroscope pushes forward, its tip moves towards the fundus of the stomach, and when the gastroscope pulls back, its tip moves away from the fundus. Thus, the gastroscope body is fixed in the upper gastric body by the stomach wall without being suspended. We called this endoscopy progression technique: Forward-Return Way (FRW). This new technique avoided the paradoxical movement in other advanced endoscopy progression skills, mainly the common U-turn. (Please refer to Figure 1 ).

All endoscopic surgeries were performed following Professor Zhi-Feng Zhao’s guidance on the FRW technique. The operators were senior endoscopists in our endoscopic treatment center with several years of experience in EFTR surgery. The patients underwent endotracheal intubation under intravenous anaesthesia before the endoscopic surgery. The anaesthesia of choice for the endoscopic surgeries of STMF was CO2 (22, 23). For the endoscopic resection of STMF, we mainly used EFTR (24).

The following data were collected and analyzed: (i) basic information of patients; (ii) whether the FRW technique was successfully performed; and (iii) the selected endoscopic method (after successful FRW). The lesion sites suitable for the FRW technique were selected according to the division of the gastric fundus and the lesion area, as shown in Figure 2 . Moreover, the success rate, results, and operation characteristics of the FRW technique were analyzed.

Manipulating performance evaluation criteria after FRW implementation ( Table 1 )

After the EFTR procedure, patients were confined to bed rest, food fasting, and water fasting for three days. The above measures and gastrointestinal decompression were performed for five days after EFTR. Parenteral nutrition was given during fasting, and continuous oral proton pump inhibitor and gastric mucosal protective agents were administered 6-8 weeks postoperatively. Patients were observed for complications of EFTR, including bleeding, perforation, and infection (25).

All patients were followed up regularly with gastroscopy examinations at 3, 6, and 12 months after the operation, and then they were checked once a year to monitor wound healing and to look for any residual or recurrent tumors.

H&E staining and immunohistochemical staining were conducted to identify the nature of the lesion and to examine any tendency to malignant transformation.

Statistical analyses were performed using IBM SPSS Statistics 25.0 (SPSS Inc., Chicago, IL, USA). Continuous data were expressed as `X ± s. Categorical data were presented as numbers or percentages (%). FRW handling performance score was expressed as X ± s.

A total of 112 cases diagnosed with STMF and treated in the Fourth Affiliated Hospital of China Medical University from 2016 to 2020 were evaluated. Ten patients qualified for the FRW technique (2 males and 8 females), aged 57.6 ± 10.07 (46-78), with a lesion diameter of 0.5-5.0cm. The FRW technique was successfully performed in 7 patients (70%) and failed in 3 patients (30%). (See Table 2 ).

Using the FRW technique, the tip of the gastroscope could access the location just below the gastric cardia in 7 out of 10 patients. All angles of the fundus of the stomach were operable in all 7 patients. Moreover, the FRW technique made accessing and observing the lesion much simpler without the need for paradoxical movement.

The FRW technique allowed clear visualization of the lesions in the entire fundus area and the mucosa just below the gastric cardia. ( Figure 3 ).

The FRW technique was mainly used for submucosal injection and lesion dissection during the endoscopic treatment of STMF. However, post-resection trauma suturing was performed in a conventional endoscopy progression.

With the FRW technique, we were able to access and visualize the edge close to the pylorus (7/7), the edge close to the gastric cardia (7/7), and the left and right edges (7/7). The FRW technique can be applied at every site, and no obvious lesion site restriction was observed.

As for the Operational Performance Evaluation of the FRW technique, it was very accessible to navigate the tip of the gastroscope tip left or right using the small adjusting knobs. However, moving the gastroscope forward and backwards still showed moderate difficulty. The gastroscope’s body makes a forward return along the gastric wall, keeping the gastroscopy’s tip stable at all angles due to the support of the gastric wall. The gastroscope was stable at every angle without the need for an assistant to support the endoscope, and the operator could manually adjust the devices in the instrument channel to complete a variety of complex adjustments, cutting, and haemostasis. Since the gastroscopy’s tip could access the lesion’s edges with stable visualization, the details of each layer below the mucosa and above the lesion were clearly observed; however, water accumulation in the lesion area had certain interference. (See Table 3 ).

The dissector can be applied parallel to the muscularis propria, forming a good inclination angle during the dissection of the STMF (as shown in the view in Figure 3B ), except for tumors at the posterior wall of the fundus. There was some resistance while progressing the instruments in the instrument channel using the FRW technique; however, this defect was the same as in the traditional U-turn technique.

The operation duration using the FRW technique was 47-122 min, and the hospital stay was 7.29 ± 0.49 (7-8) days. Postoperative histopathological evaluation revealed very low-risk stromal tumors in 8 cases and low-risk stromal tumors in 2 cases. No complications were observed, and all patients recovered well after the operation. No residual or recurrent tumors, metastasis, or death occurred during the postoperative follow-ups.

Our team discovered the FRW technique during an STMF procedure. The present study summarises the operation method and success rate of the FRW technique based on practice and analysis of the results. Unlike the traditional progression method, in which the gastroscope’s tip usually reaches the antrum first, the FRW technique uses the resistance of the posterior wall of the upper gastric body (near the junction with the fundus) and the greater curvature as the supporting force to progress the gastroscope along the arc from the posterior wall of the gastric body to the angular incisure, making a forward-return along the gastric wall. This route allows the gastroscope to access the fundus of the stomach and even access just below the gastric cardia (as shown in the view in Figure 3B ). The fundus of the stomach can then be handled from all angles. As a result, there is no need for paradoxical movement, making it much easier for the operators to access the lesion for more visualization and handling during operation.

Additionally, when observing the fundus of the stomach using the FRW technique, we could clearly visualise the mucosa just below the gastric cardia. The gastroscope’s tip had easy access to the lesion at a close distance, which can significantly benefit operating on lesions below the gastric cardia.

The fundus of the stomach is considered a challenging area for ESD (17), ESE, and EFTR (18, 19). Moreover, the ESD technique is complex and prone to perforation (26) because it intentionally destroys the full thickness of the gastric wall, causing perforation to remove the lesion then sutures the wound by nylon suture and or haemostatic clips (27). EFTR is one of the current endoscopic treatments for STMF is EFTR (16), which is why EFTR surgery was predominant in our surgeries (7/7).

Endoscopic operation on the gastric fundus is challenging, mainly due to the need for the traditional endoscopic progression method to be bent backwards (U-turn) to access and observe the fundus of the stomach completely. As a result, the direction of endoscopy body movement is opposite to the direction of observation (paradoxical movement), and the endoscopy body gets suspended, limiting the operating space for the fundus of the stomach (17). Furthermore, when the endoscope’s body is suspended using the traditional U-turn technique, the forward injection force becomes greatly reduced, and the injection needle is perpendicular to the muscularis propria (as shown in the view in Figure 3A ), making it difficult to find the submucosal space. Following the current commonly used technique, we found that endoscopic treatment of STMF lesions in the middle of the fornix and anterior wall of the fornix was extremely difficult, especially the anterior wall of the fornix at the side of the greater curvature. Thus, a better endoscopic technique was needed to observe, deliver surgical instruments, and operate on these difficult tumor sites.

The FRW technique solves the problem of avoiding paradoxical movement. Furthermore, the stability score of the endoscopy body was 2 ± 0 points, which was excellent, demonstrating that this technique can achieve non-suspension of the endoscopy body and ensure its stability. Additionally, the handling of the endoscopy body and devices in the channel would not be significantly affected by respiratory and heartbeat beats. Finally, the FRW technique allowed easy access to the lesion and its observation.

The ease of device entry and exit using the FRW technique in the 7 patients was 1 ± 0 points, suggesting that the FRW technique had a certain impact on device entry and exit. The injection needle and mucosotomy knife were not affected (2 ± 0 points), while hemostatic forceps and hemostatic clips were significantly affected (0.71 ± 0.49 points). In 2 patients, the hemostatic forceps and hemostatic clips could not pass through the curved channel, and we had to complete the pre-insertion of the device by unbending the endoscopy body and then bending it again to reach the lesion. The difficulty of the passage of the device is mainly related to the long rigid part at the front end of the device and the large bending at the channel.

The flexibility score of the endoscopy body when using the FRW technique was 1.43 ± 0.54 points, but the movement coordination and comfort of the endoscopy body were still somewhat hindered. At present, the synchronization of front and back movement of the endoscopy body is relatively unaffected, but the accuracy is slightly affected. Furthermore, the small knob swing affects the left-right movement and angle adjustment of the endoscopy body.

The success rate of the FRW technique was 70% in our clinical practice. Currently, the FRW technique can only be performed while patients are in the supine position and under general anaesthesia with endotracheal intubation. In the three cases where the FRW technique failed, we tried gastric cavity morphology after various gas volume adjustments and postural changes, including left lateral decubitus and supine positions, but FRW was still not applicable. Whether the failure cases were due to specific characteristics of the gastric fundus or other factors remains to be explored through a larger sample size. The present study had some limitations due to the small sample of STMF patients who qualified for the FRW technique. Therefore, more large-scale observational studies are needed to further validate the FRW technique. Additionally, this study was not a randomised clinical trial with a large sample size since GIST is an uncommon disease. In future, we plan to conduct clinical trials for the FRW technique to confirm its reliability.