We analyzed 1319 patients with 1443 lesions who underwent ESD for gastric tumors at Yonsei University Health Care Center between January 2008 and December 2010.

Endoscopy with standard upper gastrointestinal endoscopes (GIF Q260 and H260, Olympus, Japan), chromoendoscopy with indigo carmine, and biopsies of the lesions with standard biopsy forceps (FB-21K-1; Olympus, Japan) were initially performed to determine the feasibility of ESD. Endoscopic ultrasonography (EUS) with radial scanning echoendoscopes (EG-3679URK, Pentax, Japan and GF-UE260, Olympus, Japan) was performed in case of carcinoma to evaluate the depth of invasion. Patients with lesions confirmed to be gastric cancer underwent abdominal computed tomography (CT) scans to determine if lymph node or distant metastasis was present. For this study, the endoscopic findings of EGC were classified as elevated (types I or IIa), flat (type IIb), depressed (types IIc, IIc+III, or IIa+IIc) or mixed (types IIa+IIb, IIb+IIc, IIa + IIc, III+IIa or III+IIb).

Gastric tumor lesions were classified according to their location in the upper, middle, and lower thirds of the stomach, and also to location in the cross-sectional circumference divided into four equal parts (anterior wall, posterior wall, lesser curvature or greater curvature).

Eligibility criteria for ESD were as follows: (1) differentiated adenocarcinoma (well- to moderately-differentiated tubular adenocarcinoma) or dysplasia confirmed histologically by forceps biopsy; (2) depth of invasion limited to the mucosa or submucosa (≤ 500 μm penetration into the submucosa) as determined by EUS; (3) lesions without ulceration, regardless of size, or 30 mm or less in size with ulceration; or (4) undifferentiated adenocarcinoma or intramucosal cancer without ulcer findings ≤ 20 mm in size[8].

All procedures were performed by an attending gastroenterologist, and five attending physicians were involved in the procedures. All ESDs were performed under conscious sedation using intravenous propofol or midazolam. Vital signs were continuously monitored during the procedure. After identifying the target lesion, marking dots were made circumferentially about 5 mm lateral to the margin of the lesion using a needle knife (KD-10Q, Olympus, Japan) or argon plasma coagulation (ERBE Elektromedizin, Germany). Epinephrine (1:10000 dilution) was then injected submucosally around the lesion, and an initial short incision was made in the mucosa with a needle knife to allow submucosal insertion of the tip of an insulation-tipped (IT) knife (KD-611L, Olympus, Japan). Circumferential mucosal cutting was performed outside the marking dots, and an additional submucosal injection was carried out. Finally, direct dissection of the submucosal layer was performed, and endoscopic hemostasis with specialized hemostatic forceps (FD-410LR, Olympus, Japan) was performed when needed.

All resected specimens were systematically sectioned at 2 mm intervals centered on the part of the lesion closet to the margin and the site of the deepest invasion. Histological assessment was based on the Vienna classification[9].

Final pathologic diagnoses were classified as low grade dysplasia (LGD), high grade dysplasia (HGD), differentiated EGC, and undifferentiated EGC.

Patient data, including patient age, gender, previous medication history, the size, number and location of lesions, procedure start and end times, endoscopic findings, pathology, and complications, were collected.

Complete resection of en bloc resected tumors was defined as the lateral and vertical margins being free of tumor cells on histologic examination. Complete resection of tumors resected in a piecemeal fashion was defined as complete removal of the entire lesion, including sufficient tumor-free margins after perfect reconstruction of all pieces.

Procedure time was defined as the time from marking to complete removal, including the time required for hemostasis. Complication data included whether a complication occurred and details regarding bleeding, perforation and other factors related to the type of complication.

To identify factors affecting the success of ESD, we analyzed lesion characteristics, procedure, and the procedure result. Analyzed lesion characteristics included the presence of ulceration, macroscopic morphology, size and location of the tumor.

Procedure results were analyzed for curability. Resection was deemed complete when removal was achieved with tumor-free lateral and vertical margins and there was no lymphovascular involvement or lymph node metastasis. Incomplete resection was defined as any resection that did not meet the curative criteria described above.

Endoscopic surveillance by esophagogastroduodenoscopy (EGD) was performed 3, 6, 12, and 24 mo after ESD for EGC to exclude local recurrence, as well as synchronous, and metachronous lesions. After 24 mo, EGD was carried out annually. Moreover, abdominal CT scans were performed every 6 mo for the first year and annually thereafter, to detect lymph node or distant metastasis. In cases with adenomas, endoscopic surveillance by EGD was scheduled for 3, 12, and 24 mo after ESD.

The data were analyzed using Pearson’s χ² test, unpaired t-test, Fisher’s exact test, and the Mann-Whitney U test. P values < 0.05 were considered significant. To identify related risk factors for complications and complete resection, predictors with P values < 0.05 in the univariate analysis were included in a backward, stepwise multiple logistic regression model. All data analyses were conducted using a statistical software package (SPSS version 18.0, Chicago, IL, United States).

During the study period, ESD was performed in 1319 patients with 1443 gastric tumors. Baseline clinicopathologic characteristics of the gastric tumors and the clinical outcomes of ESD are shown in Table 1. Mean age was 63.0 ± 9.4 years. The lesions consisted of 733 (50.8%) EGCs and 710 (49.2%) dysplastic lesions. Submucosal invasion occurred in 7.3% of cases. Mixed-type endoscopic morphology was the most common (63.4%). With respect to size and location, tumors less than 20 mm in size (71.7%), those located in the lower third (85.4%) and those located in the lesser curvature (33.3%) were most common. The mean tumor size was 15.72 ± 8.81 mm. The mean procedure time was 61.8 ± 47.0 min. The complete resection rate was 89% (1287/1443), and the en bloc resection rate was 91.3% (Table 1). The post-ESD bleeding rate was 4.3%, and the perforation rate was 2.7%. Most cases of bleeding (60/63) were treated by endoscopic hemostasis such as hemoclipping, argon plasma coagulation or epinephrine injection. Two cases were treated by angiographic embolization. Only one case required surgery for bleeding control.

Around half of all perforation cases (20/39) were minute or micro-perforations, while the remaining ones were overt perforations. Only two such cases required surgery. All other cases were treated by conservative care. There was no mortality in the present study.

We compared the clinical outcomes of ESD in relation to detailed tumor location. Upon division into upper third and other lesions, the upper third lesion group had significantly higher percentages of incomplete resections (19.4% vs 10.2%, P = 0.005) and piecemeal resections (15.3% vs 8.2%, P = 0.015) compared with other tumor locations. Additionally, upper third lesions required a longer procedure time (90.51 min vs 59.71 min, P < 0.001) and were associated with a higher perforation rate (9.2% vs 2.2%) (Table 2). There was no significant difference in the frequency of post-ESD bleeding.

After dividing location according to posterior wall and other lesions, the posterior wall lesion group had a significantly longer procedure time (69.41 min vs 58.84 min, P < 0.001) and higher rates of incomplete resections (14.8% vs 9.2%, P = 0.002) and piecemeal resections (11.6% vs 7.5%, P = 0.013) than lesions in other locations (Table 3). There was no significant difference in the frequency of post-ESD bleeding or perforation between the two groups.

We analyzed the factors associated with complete resection complications of ESD such as perforation or bleeding. In univariate analyses, lesion size larger than 20 mm, upper third location, posterior wall location, carcinoma and procedure time longer than 60 min were significantly related to incomplete resection. In multivariate analysis, lesion size larger than 20 mm, posterior wall location, carcinoma and procedure time longer than 60 min were significantly related to incomplete resection (Table 4).

In addition, univariate predictors of perforation were lesion size larger than 20 mm, upper third location, procedure time longer than 60 min and piecemeal resection. In multivariate analysis, upper third location, procedure time longer than 60 min and piecemeal resection were statistically significantly related to perforation (Table 5).

Moreover, univariate predictors of post-ESD bleeding were lesion size larger than 20 mm, procedure time longer than 60 min and piecemeal resection. In multivariate analysis, procedure time longer than 60 min and piecemeal resection were statistically significantly related to post-ESD bleeding (Table 6).

Advances in diagnostic technology and the increasing prevalence of screening programs have increased the rate of early gastric cancer (EGC) detection. EGCs that are confined to the mucosa and lack lymph node metastasis can be cured by endoscopic resection, such as endoscopic mucosal resection and endoscopic submucosal dissection (ESD).

Previous studies analyzed tumor location which is anatomically divided into three portions (upper, mid, and lower thirds) as a factor related to the clinical outcomes of ESD.

In this study, the perforation rate was significantly associated with upper third location. This may be due to the fact that lesions at lower third are easily approached and manipulated by endoscopy, and thus it is technically easier to perform ESD and there is a lower chance of applying sufficient tension on the gastric wall to cause perforation.

ESD for gastric tumor is an effective and safe therapy. However, endoscopists should recognize the need for more advanced endoscopic techniques when performing ESD for lesions located in the upper third or posterior wall of the stomach in order to decrease the rate of serious complications and improve clinical outcomes.

This is an interesting study regarding on the technical aspect of gastric ESD, focused on the tumor location. The number of the subjects is large and the analysis is simple with clear results.