The Korean ESD registry database is an online registry created in 2015. It is an ongoing project collecting ESD or EMR data from multiple centers representing different geographical areas in Korea (https://esd.gie.or.kr) and is under quality control of the Korean Society of Gastrointestinal Endoscopy. This registry contains clinical information, endoscopic findings, therapeutic outcomes, and follow-up data, including recurrence and survival related to ER of gastric neoplasms. Cases with ER of EGCs with undifferentiated-type histology were identified and reviewed. Clinical, endoscopic, and histologic findings before and after ER were evaluated retrospectively from the original database.

ESD or EMR procedures were performed using standard techniques (marking, submucosal injection, incision, submucosal dissection/injection, and snaring), although different types of instruments, including endoscopes, electrosurgical units, sedative drugs, and endoscopic knives were used. ER of small (less than 2 cm) non-ulcerative intramucosal gastric cancer with undifferentiated-type histology was defined as expanded indication or criteria, according to Korean guidelines [5]. ER of EGCs with undifferentiated-type histology that did not meet this standard was defined as beyond expanded indication or criteria. Pre-ER indication refers to the judgment of the endoscopist as to whether the lesion is candidate for ER or not, according to the guidelines [5]. Post-ER criteria is the result of histologic evaluation as to whether the resected lesion is included in the expanded or beyond expanded criteria [5].

Follow-up endoscopic surveillance was performed every 3 to 6 months for 1–2 years and every 6 months for the next 3–5 years, according to the national comprehensive cancer network (NCCN) guidelines [8]. Abdominal and pelvic computed tomography using a contrast, and chest X-ray were performed every 6 months for 5 years.

Immediate therapeutic outcomes, including en bloc resection, complete resection, and curative resection rates, and long-term outcomes, including recurrence and mortality rates, were extracted and analyzed. Therapeutic outcomes, divided by pre-ER indication and post-ER criteria, were evaluated to identify the suitability of current indications.

En bloc resection was defined as complete removal of cancer in a single piece without fragmentation. Complete resection was defined as removal of cancer with no neoplastic components at the lateral or vertical margins and without lymphovascular invasion on microscopic examination. Curative resection was defined as removal of cancer with 20 mm or smaller intramucosal lesions without ulceration, neoplastic components at the lateral or vertical margins, and lymphovascular invasion [3]. Recurrence was defined as reappearance of cancer at the site of the lesion (local recurrence) or synchronous, metachronous, or distant metastatic lesions.

Continuous variables were expressed as the median and interquartile range (IQR) because they were not normally distributed. Categorical variables were expressed as numbers and percentages. Mann–Whitney and Fisher’s exact tests were used to compare two variables. In this study, a p-value <0.05 (2-tailed) was adopted as the threshold of statistical significance for all tests. The analyses were performed using SPSS version 20.0 (SPSS Inc., Chicago, IL, USA).

This study was conducted according to the principles expressed in the Declaration of Helsinki, and approved by the Institutional Review Board of Chuncheon Sacred Heart Hospital (2016-87).

From 2006 to 2015, 275 EGCs with undifferentiated-type histology from 275 patients were identified. Of the total lesions, 186 were PDC (67.6%) and 89 were SRC (32.4%). The median age of the enrolled population was 69 years (IQR, 60–76 years), with a male predominance (165 men, 60%). The antrum was the most frequent location of lesions (47.6%), followed by the lower body (30.2%), upper body and cardia (12%), and midbody (10.2%). The macroscopic types of the lesions were elevated (36.7%), flat (35.6%), and depressed (27.6%). The median pre-ER endoscopic size measurement of the lesions was 25 mm (IQR, 15–40 mm) and the median post-ER histologic size measurement was 18 mm (IQR, 11–26 mm). With regard to the depth of invasion, 64% were mucosal lesions and 36% exhibited submucosal invasion. The pre-ER histology was variable and included PDC (32.4%), SRC (21.5%), moderately differentiated adenocarcinoma (20%), well-differentiated adenocarcinoma (11.6%), low-grade dysplasia (5.5%), high-grade dysplasia (4.7%), and atypical cells or unknown (4.4%). The vast majority of lesions were resected using the ESD method (90.9%), and 9.1% of lesions were resected using the EMR method. Helicobacter pylori were detected in 40% of the enrolled population.

In terms of the pre-ER indications, only 82 lesions (29.8%) were included in the expanded indication group, and 193 lesions (70.2%) were included in the beyond expanded indication group. In the post-ER histologic criteria, 108 lesions (39.3%) and 167 lesions (60.7%) fulfilled the expanded and beyond expanded criteria, respectively. There was discrepancy between pre-ER indications and post-ER criteria in 36.7% (101/275) of all lesions. The clinical and endoscopic characteristics of the enrolled population are summarized in Table 1.

The immediate therapeutic outcomes were as follows: en bloc resection rate: 92.4%; complete resection rate: 80%; and curative resection rate: 36.4%. Lesions with pre-ER expanded indication were associated with better therapeutic outcomes than those of lesions with pre-ER beyond expanded indication, although significant differences between these two groups were detected only with regard to the curative resection rate (48.8% vs. 31.1%, p=0.006). The submucosal invasion rate was high regardless of indications, which was 34.1% in lesions with pre-ER expanded indication and 36.8% in lesions with pre-ER beyond expanded indication. The lymphovascular invasion rates were 7.3% and 10.4% in lesions with pre-ER expanded indication and lesions with pre-ER beyond expanded indication, respectively (p=0.51). The procedure-related complication rates were 0.7% for delayed bleeding and 1.8% for gastric perforation. These rates were not significantly different between lesions with pre-ER expanded indication and beyond expanded indication. The median procedure time was 50 minutes (IQR, 33–76 minutes), and lesions with pre-ER beyond expanded indication exhibited longer procedure time than that exhibited by lesions with pre-ER expanded indication (60 minutes vs. 45 minutes, p=0.02). Detailed immediate therapeutic outcomes according to the pre-ER indication are described in Table 2.

Compared to lesions with post-ER beyond expanded criteria, those with post-ER expanded criteria were associated with better therapeutic outcomes. The en bloc resection rate was not significantly different between these two groups (95.4% vs. 90.4%, p=0.16). However, complete resection rate was higher in the lesions with expanded criteria than in lesions with beyond expanded criteria (91.7% vs. 72.3%, p<0.001). Of the lesions with post-ER beyond expanded criteria, 59.6% exhibited submucosal invasion. The lymphovascular invasion rate was 3.7% in lesions with post-ER expanded criteria and 13.3% in lesions with post-ER beyond expanded criteria (p=0.01). Procedure-related complication rates were not significantly different between lesions with post-ER expanded criteria and beyond expanded criteria. Lesions with post-ER beyond expanded criteria had a longer procedure time than that of lesions with post-ER expanded criteria (59 minutes vs. 41 minutes, p=0.002). Detailed immediate therapeutic outcomes according to the post-ER criteria are described in Table 3.

There was no difference in immediate outcomes between PDC and SRC, including en bloc resection, complete resection, and curative resection rate. However, ER of PDC was associated with vertical margin involvement (9.7% vs. 3.4%) while ER of SRC was associated with lateral margin involvement (6.7% vs. 4.3%), although these associations were not statistically significant (p=0.21). ER of PDC rather than SRC was more strongly associated with submucosal invasion (41.9% vs. 23.6%, p=0.003). The rate of lymphovascular invasion was also higher in ER of PDC than of SRC, although this difference was statistically insignificant (11.3% vs. 5.6%, p=0.19). The rate of procedure-related complications including delayed bleeding and gastric perforation was not significantly different between ER of PDC and SRC (p=0.54 & p=0.70, respectively). The procedure time was not significantly different between ER of PDC and SRC (50 minutes vs. 50 minutes, p=0.94). Detailed immediate therapeutic outcomes according to the histology are described in Table 4.

Fig. 1 shows the treatment flow of ER of EGC with undifferentiated-type histology. With regard to the long-term outcomes, the recurrence rate was 10.2% and median recurrence period was 13.6 months (IQR, 2.4–47.7 months) after ER. A total of 5 (1.8%) deaths were observed in the overall population during the observational period (median follow up: 3.96 years), and these were not related to cancer (Table 5). All of the recurrence cases were treated by repeated ESD or operation, and there was no recurrence at the last follow-up (Fig. 1).

In terms of the recurrence rate, there was no difference between lesions with curative resection and non-curative resection (p=0.28). This rate was also not different between PDC and SRC (p=0.11). Although lesions with pre-ER expanded indication showed higher recurrence rates than those with pre-ER beyond expanded indication (14.6% vs. 8.3%, p=0.02), this rate was not statistically significant (9.2% vs. 10.8%, p=0.21) (Table 5).

In terms of the mortality rate, there was no difference between lesions with curative resection and non-curative resection (p=0.32). This rate was not different between PDC and SRC (p=0.43). Mortality rates according to the pre-ER indication and post-ER criteria showed no significant differences (p=0.65 and p=0.30, respectively) (Table 5).