Clinical efficacy of endoscopic resection for subepithelial tumors in the esophagogastric junction and gastric cardia: an observational study

Patients

A total of 26 consecutive patients with SETs at the EGJ and cardia underwent endoscopic resection at the Presbyterian Medical Center between November 2012 and May 2023. Indications for resection included increased tumor volume and size (tumors exceeding 2 cm). All patients underwent endoscopic ultrasonography (EUS) prior to endoscopic resection. None of the patients showed metastases on computed tomography.

Procedures

Overall, 27 SETs were resected via endoscopic muscularis dissection (EMD), ESSD, or submucosal tunneling endoscopic resection (STER) (Figs. 1, 2). Conscious sedation was used for most of these procedures. General anesthesia was necessary for two patients in whom the procedure time was expected to be long. A single skilled endoscopist performed all procedures using the following instruments: gastroscope (GIF-HQ290 or GIF-H260; Olympus), hook knife (KD-620LR; Olympus), I-type knife (FM-EK0003-2; Finemedix), ITknife2 (KD-611L; Olympus), and electrical cutting device (ERBE ICC 200; ERBE).

Fig. 1.

Endoscopic subserosal dissection for a gastrointestinal stromal tumor. (A) A subepithelial tumor identified in the gastric cardia. (B,C) The incision extending into the muscularis propria layer. (C) The subserosal layer is seen clearly. (D) The wound surface is shown as an intact serosa after tumor removal. (E) Incision closure with clips. (F) Macroscopic appearance of the resected tumor.

Fig. 2.

Submucosal tunneling endoscopic resection for a leiomyoma in the esophagogastric junction (EGJ). (A) Endoscopic view of a subepithelial tumor (SET) in the EGJ and gastric cardia. (B) Establishment of a submucosal tunnel. (C) Exposure of the SET. (D) Macroscopic appearance of the resected tumor.

Outcome measures

The rates of en bloc resection, complete resection, and recurrence were calculated to evaluate the efficacy of the endoscopic resections. Complete resection was defined as an en bloc resection without residual endoscopic lesions. To evaluate safety, complication rates were recorded. Complications included post-procedural bleeding, peritonitis, and infection.

Statistical analysis

Baseline characteristics and treatment outcomes were compared between the groups. Student’s t-test, Mann-Whitney U-test, and Fisher’s exact test were used. All data were analyzed using IBM SPSS ver. 22.0 (IBM Corp.), with statistical significance set at p<0.05 (two-tailed).

Ethical statements

The Institutional Review Board of the Presbyterian Medical Center approved this study (approval number: 2024-01-003). Informed consent was obtained from all the patients.

General characteristics

This study included 26 consecutive patients with an average age of 54.3 years (range, 25–74 years). One patient had two SETs, while the other 25 had one tumor each. Of the 27 gastric SETs, 3 and 24 were located in the EGJ and cardia, respectively. The mean tumor size was 21 mm (range, 5–52 mm). Pathological examination confirmed the presence of 4 GISTs and 23 leiomyomas (Table 1).

Table 1.

Baseline characteristics and outcomes of endoscopic resection

Treatment outcomes

The mean procedure time was 35.4 (range, 9–65) minutes, and the mean hospitalization duration was 4.2 days (range, 3–7 days). Endoscopic resection achieved an en bloc resection rate of 100% (27/27) and a complete resection rate of 88.9% (24/27). No postoperative bleeding or infection was observed, and no recurrence was observed during follow-up (range, 3–24 months). Residual lesions were identified in three patients (11.1%). Cases 1, 5, and 6 underwent incomplete resections. All three cases were leiomyomas and are currently being followed up for 10, 6, and 7 years, respectively. All cases showed slow growth for 4 to 5 years after resection; however, no further growth was observed thereafter, and the patients have not required further treatment other than regular follow-up (Tables 2, 3).

Table 2.

Treatment outcomes

Table 3.

Clinical characteristics and treatment outcomes

Differences according to resection techniques

Overall, 17 of 27 gastric SETs were treated using EMD. Additionally, 9 SETs were removed using ESSD, and 1 SET was resected using STER (Table 2). To compare the outcomes according to the resection technique, the patients were divided into two categories. Group A: resection without penetration of the MP layer (i.e., EMD) and group B: patients underwent resection beyond the MP layer (i.e., ESSD or STER) (Table 4). No significant differences in baseline characteristics were observed between the two groups. Complete resection was achieved in 14 (82.5%) patients in group A and 10 (100%) patients in group B. Residual lesions were identified in three patients (17.7%) in the EMD group and none in the other group (Table 4).

Table 4.

Comparison of outcomes according to resection techniques

Comparisons based on different resection periods

To compare outcomes according to the endoscopist’s experience and skill level, patients were divided into two categories. Group I: resections performed before 2020 and Group II: resections performed 2020 and after. There were no significant differences in general characteristics between the two groups (Table 5). Complete resection was achieved in 10 (76.9%) patients in group I and 14 (100.0%) patients in group II (Table 5). Residual lesions were identified in three patients (23.1%) in group I and none in group II. No significant differences in tumor size (group I vs. group II, 17.9 vs. 23.9 mm; p=0.186) and operative time (group I vs. group II, 33.9 vs. 36.8 minutes; p=0.676) were observed between the two groups. The correlation coefficients between the tumor size and operative time were 0.795 and 0.547 for groups I and II, respectively (Table 6).

Table 5.

Comparison of outcomes during different periods

Table 6.

Comparison of tumor size and operative time between groups I and II

Endoscopic techniques for the resection of gastric SETs

Endoscopic submucosal dissection (ESD) techniques for SETs are referred to in various terms.15 In our study, we used the term “endoscopic muscularis dissection”. For cases of SETs with a narrow muscular connection identified on EUS, ESD is expected to be used effectively with minimal damage to the muscle layer.3,16 Complete resection rates of conventional ESD for tumors from the MP layer have been inconsistent among studies, varying from 68.2% to 92.4%.17 In a large-scale study by He et al.,18 the complete resection rate of SETs by ESD was 92%, and the perforation rate was 14%. However, only 13.1% of these tumors were located in the gastric cardia. As mentioned, SETs of MP origin located at the EGJ and gastric cardia are difficult to completely resect using endoscopy, and several advanced techniques have been attempted.2,13,19-22

EFTR is emerging as a novel procedure for resecting SETs, especially when dealing with SETs embedded within the MP layer.9,13,19,20 However, EFTR has limited applications depending on the location and size of the tumor. As EFTR creates an artificial wall defect, there may be a risk of shedding tumor cells, and complete closure is important to prevent peritonitis.3

In contrast, STER is a method of forming a tunnel that has several advantages, such as mucosal integrity and easy closure of the mucosal defect.20 However, this technique requires sufficient space in the submucosal tunnel with a direct view, and it was performed in only one case that conformed to the conditions in our study. In contrast, ESSD can be performed with better visibility than STER; hence, it is effectively used for tumors of the lesser curvature or gastric cardia.10 ESSD is a new countermeasure for the resection of gastrointestinal SETs with exophytic growth.22

In the present study, among the 27 SETs located in the gastric cardia and EGJ, nine were resected with ESSD, and the complete resection rate was 100% without complications (Table 3). Because of the relatively thick subserosa of the gastric cardia and its lesser side, ESSD can be used effectively and safely for the resection of deeply embedded tumors in these areas.10,11 Considering the limitations of EFTR and STER, we suggest that ESSD should be used as a new alternative for the treatment of SETs around the gastric cardia.

Choice of resection techniques

The endoscopic resection method was planned prior to starting the procedure according to the location and EUS findings of the target SETs. (1) EMD was planned if the tumor growth pattern was endophytic and had a narrow muscular connection. (2) If the tumor growth pattern was exophytic, subserosal, and/or large, ESSD was planned for complete resection of the tumor. Among the tumors located at the EGJ, STER was performed in only one case in which sufficient submucosal tunnel space could be secured (Table 2). The complete resection was achieved in 14 (82.5%) patients with EMD, whereas it was achieved in 100.0% in patients with ESSD and STER. We found that more invasive resection techniques, such as ESSD or STER, had higher complete resection rates than conventional ESD. Additionally, the lack of experience with endoscopic resection contributes to incomplete resection.

Choosing the optimal endoscopic treatment strategy for SETs remains challenging. Chun et al.16 reported that ESD is indicated for small SETs (size ≤20 mm). In the current study, the mean resected tumor size using EMD was 17.9 mm, which was smaller than that in the ESSD or STER groups (26.3 mm). This finding is consistent with previous studies, and shows that ESD or EMD can be used effectively and safely even for SETs at the EGJ and gastric cardia. Among the patients who underwent EMD, residual lesions were identified in three cases (17.6%), but none of the patients who underwent complete resection experienced recurrence. Taken together, EMD can be effectively and safely used for the resection of small-sized (≤20 mm) SETs, while ESSD should be considered for the resection of large-sized (≥20 mm) deeply embedded tumors located in the EGJ and gastric cardia.

Endoscopist’s experience and prerequisites for the procedure

In patient group I, where the operator's experience was relatively limited, a complete resection rate of 76.9% (10/13) and three cases of residual lesions were identified. However, complete resection was achieved in all patients since 2020 (100.0%, 14/14).

Endoscopic manipulation for resection of tumors at the EGJ and gastric cardia is technically difficult. These procedures require manipulation of the deep layers of the stomach wall, particularly in anatomically complex areas, such as the EGJ and cardia. Therefore, these procedures should only be performed by endoscopists with extensive experience in ESD and therapeutic endoscopy. It is desirable to first observe the procedure by an experienced doctor, and then begin the procedure under their guidance. Han et al.10 mentioned that an operator with more than 100 gastric ESD experiences may start ESSD, which was used in 9 patients in this study.

In the present study, one skilled operator with experience of more than 2000 gastric ESD cases performed all procedures. The operator showed a higher complete resection rate after accumulating experience in more than ten cases; however, there were no specific benchmarks for the learning curve. Meanwhile, the operator accumulated experience with other advanced endoscopic techniques, which may have had an impact on improving their skill levels. Therefore, it is difficult to define the number of cases that would lead to proficiency in the resection of SETs at the EGJ and cardia at approximately 10.

The effect of endoscopist experience on the relationship between tumor size and procedure time was evaluated using Pearson’s correlation coefficient. We found that the correlation coefficient in group II, in which patients who had endoscopic resections since 2020, was 0.547, which was lower than that in group I (0.795). Generally, the procedure time increased as the tumor size increased (Pearson’s correlation coefficient approaches 1). However, the correlation coefficient was lower in group II than in group I. This may be attributed to the operator becoming more skilled, and thus, the procedure time did not increase significantly even when the tumor size increased. This reflects an increase in the skills of the endoscopist.

The endoscopic resection of SETs in the EGJ and gastric cardia requires highly advanced techniques. For safe and effective treatment, it is necessary to perform the procedure at a large specialized center with multidisciplinary support capable of performing emergency surgery in cases of complications, such as perforation. Although a standardized method has not yet been established, many technological developments have been made recently to increase complete resection rates and reduce complications. For ESD to be used more safely and effectively as a standard treatment, operators must have a comprehensive understanding of the anatomical structure of the stomach wall and substantial procedural experience.

Limitations

The present study has several limitations. This is a retrospective study. As such, selection bias may be unavoidable. The sample size was small, which may have affected the establishment of statistical significance. In the future, large-scale population-based studies should be conducted.

In conclusion, endoscopic resection can be used effectively for SETs in the EGJ and gastric cardia. It is a promising treatment alternative to surgery for tumors located in the EGJ and gastric cardia.