The detection rate of early gastric cancer (EGC) has increased significantly owing to the widespread implementation of national endoscopy screening programs, particularly in East Asia. Therefore, the treatment paradigm for these lesions has shifted from surgical gastrectomy to endoscopic resection. Endoscopic submucosal dissection (ESD) has replaced conventional endoscopic mucosal resection (EMR) as the preferred treatment modality for early stage lesions with a negligible risk of lymph node metastasis (LNM). ESD offers the distinct advantage of achieving en bloc resection regardless of lesion size, which allows for precise histological assessment and significantly reduces the risk of local recurrence compared to EMR [1].

However, ESD is a technically demanding procedure associated with a longer procedure time and a higher risk of adverse events such as bleeding and perforation. Moreover, endoscopists increasingly encounter elderly patients with multiple comorbidities, necessitating a careful balance between the risk of thromboembolism and procedural hemorrhage. This review discusses the latest global guidelines concerning the indications and clinical outcomes of ESD for EGC, with particular emphasis on expanded indications, including undifferentiated-type histology and submucosal invasion. It summarizes practical strategies for overcoming technical challenges using various traction methods and details the management of complications, including delayed adverse events, with a focus on recent updates in antithrombotic guidelines. While standard indications ensure safety, the expansion to “beyond standard” criteria allows for organ preservation in a wider patient group, provided that strict risk stratification is applied.

The indications for ESD have evolved from classic “absolute indications” to “expanded indications” based on accumulating evidence that the risk of LNM is negligible under specific conditions. Table 1 compares the current indications across guidelines from the Korean Gastric Cancer Association (KGCA 2024) [1], Japan Gastroenterological Endoscopy Society (JGES 2021) [2], and European Society of Gastrointestinal Endoscopy especially the Management of Precancerous Conditions and Lesions in the Stomach (MAPS III 2025) [3].

All three guidelines agree that differentiated intramucosal cancer without ulceration is an absolute or recommended indication for ESD, regardless of tumor size. Similarly, differentiated intramucosal cancer with ulceration is considered an absolute indication if the tumor size is ≤3 cm. The Korean guidelines specifically use the term “standard indications” for well or moderately differentiated tubular or papillary EGCs measuring ≤2 cm without ulceration and confined to the mucosa [1].

Because the definition of “standard indication” differs among regional guidelines, the term “beyond standard indication” in this review refers to lesions that are outside the Korean standard indication, but are still considered appropriate for ESD under the expanded or conditional criteria in the Japanese and European guidelines. These lesions correspond to what has traditionally been termed “expanded indications” in the Japanese and Western literature. These expanded criteria have been validated through large-scale prospective studies, particularly the Japanese JCOG0607 trial, which demonstrated that selected lesions meeting expanded criteria have excellent long-term outcomes comparable to surgery [1,2].

For differentiated-type EGC, the Korean and Japanese guidelines classify the following expanded indications: 1) differentiated intramucosal cancer >2 cm without ulceration (any size in Japanese guidelines); 2) differentiated intramucosal cancer ≤3 cm with ulceration.

However, an important distinction exists for larger ulcerated lesions: the Japanese guidelines classify differentiated intramucosal cancer >3 cm with ulceration as a “relative indication” rather than an expanded indication, reflecting the need for more cautious patient selection in these cases [2]. The European MAPS III guidelines recommend ESD for differentiated dysplastic or intramucosal lesions of any size if not ulcerated, and ≤30 mm if ulcerated [3].

Undifferentiated-type EGC, including poorly differentiated adenocarcinoma and signet ring cell carcinoma, represents a particularly challenging subset of endoscopic management because of the historically higher rates of LNM. However, recent evidence from the Japanese JCOG1009/1010 trials has demonstrated that highly selected undifferentiated-type lesions (≤2 cm, intramucosal, without ulceration) can be safely treated with ESD [4].

The Korean guideline states that after pathological review, differentiated-type cancer with minute submucosal invasion (≤500 μm) measuring ≤3 cm may be considered for curative resection if en bloc R0 resection is achieved with no lymphovascular invasion and negative vertical margin [1]. The Japanese guidelines classify this as an expanded indication [2], whereas MAPS III states that it “can be considered” for differentiated lesions ≤30 mm [3].

This difference reflects the need for rigorous pathological evaluation to ensure curative resection, particularly in undifferentiated cases. Additionally, while the Japanese guidelines explicitly include local recurrence after EMR/ESD or ESD as an indication, the Korean guidelines generally treat it as a relative indication because of the technical difficulties associated with severe fibrosis.

Successful ESD requires overcoming technical challenges, primarily severe fibrosis and difficult anatomical locations such as the cardia or pylorus. In cases of ulcer scars, the submucosal layer is often thin and fibrotic, making dissection hazardous. The use of high-viscosity injection solutions such as sodium hyaluronate is essential in these scenarios to maintain adequate submucosal elevation.

Gravity often hinders visualization of the dissection plane and increases the risk of perforation. To address this problem, various traction techniques have been developed. Table 2 summarizes the available traction methods and their advantages, disadvantages, and optimal indications.

The safety and efficacy of ESD have been validated globally in numerous large-scale studies. Table 3 summarizes the key findings from both Eastern and Western centers.

In Korea, a prospective multicenter cohort study by Kim et al. [9] reported an en bloc resection rate of 97.4% and a 5-year overall survival rate of 96.6%, confirming the long-term safety of the procedure. In an aging society, outcomes for the elderly are of particular concern. Kim et al. [10] analyzed outcomes in elderly patients aged 75 years and older, finding that while the rate of adverse events was slightly more frequent in older patients, the curative resection rate remained high, justifying the application of ESD in this population.

Although ESD originated in East Asia, it has since been successfully adopted in Western countries. Recent studies from the USA by Yang et al. [11] demonstrated that ESD is feasible in Western practice, with high technical success rates. Studies from Germany by Probst et al. [12] demonstrated that ESD is feasible in Western practice, with en bloc rates exceeding 94%. Recent analyses have also suggested that ESD achieves overall survival comparable to surgery, with significantly fewer ad-verse events [13,14].

Bleeding is the most common adverse event associated with ESD. Although intraprocedural bleeding is managed with electrocoagulation, delayed bleeding requires preventive strategies, particularly the management of antithrombotic agents. Table 4 outlines practical strategies based on the latest consensus from gastroenterology and cardiology societies. For patients taking aspirin, monotherapy may be continued even for high-risk procedures such as ESD, as bleeding risk must be balanced against cardiovascular risk. For ultrahigh-risk cases, withholding aspirin for 5–7 days may be considered during cardiology consultations [17,18]. In patients with coronary stents undergoing dual antiplatelet therapy, discontinuation of both agents is contraindicated because of the risk of stent thrombosis. The standard approach is to withhold the P2Y12 inhibitor (clopidogrel for 5 days, prasugrel for 7 days) perioperatively while continuing aspirin monotherapy [17-19]. A significant paradigm shift occurred with the use of direct oral anticoagulants (DOACs). Recent guidelines, including the 2022 American College of Gastroenterology–Canadian Association of Gastroenterology guidelines and the 2024 International Digestive Endoscopy Network consensus, emphasize that heparin bridging is no longer recommended for DOACs because it increases bleeding risk without providing a significant thromboembolic benefit [17,18].

Drug-specific DOAC management before ESD (based on the Korean IDEN 2024 consensus) [18]: 1) apixaban, rivaroxaban, edoxaban: discontinue 2 days before procedure (D-2); 2) dabigatran with normal renal function (creatinine clearance ≥50 mL/min): discontinue 2 days before procedure; 3) dabigatran in patients with renal impairment (creatinine clearance <50 mL/min) is discontinued 5 days before the procedure due to 80% renal elimination; and 4) edoxaban: extended discontinuation (3 days) may be required if renal function deteriorates, as 50% of patients undergo renal excretion.

All DOACs should be withheld for at least 48 h before highrisk procedures such as ESD based on their approximate 12-hour half-life. DOACs can be resumed 24–48 hours post-procedure if adequate hemostasis is achieved [18].

The management pathway after ESD is determined based on the curability of the resection. Patients who achieve curative resection should undergo annual surveillance endoscopy to detect metachronous gastric cancer, the incidence of which is approximately 9.5% at 5 years following endoscopic resection [21]. Furthermore, Helicobacter pylori eradication should be performed in H. pylori-positive patients because it significantly reduces the risk of metachronous recurrence [1].

If resection is noncurative, the management strategy depends on the specific reason for non-cure and the risk of LNM (Fig. 1). In cases in which non-curativity is solely due to a positive horizontal margin in a differentiated lesion, the risk of LNM is minimal. Therefore, additional endoscopic treatments, such as re-ESD, argon plasma coagulation, or close observation, may be considered. Conversely, for patients with risk factors for LNM, such as deep submucosal or lymphovascular invasion, additional surgical resection is the standard for care.

However, for elderly patients or those who represent poor surgical candidates, the eCura system can be utilized to stratify patients and determine a potential observation strategy [22]. It is important to note that the eCura system is a Japanese-derived risk stratification tool and is not explicitly adopted as the standard algorithm in the KGCA guidelines (2022/2024) or European MAPS III guidelines [1,3].

The eCura system assigns points based on: 1) tumor size >30 mm (1 point); 2) positive vertical margin (1 point); 3) submucosal invasion ≥500 μm (SM2) (1 points); and 4) lymphovascular invasion (3 points).

Risk categories: 1) low risk (eCura 0–1): LNM rate <3%, observation may be considered; 2) intermediate risk (eCura 2–4): LNM rate 5%–10%, multidisciplinary discussion required; and 3) high risk (eCura ≥5): LNM rate >15%, strong recommendation for surgery.

For undifferentiated-type EGC, the modified eCuraU system has been validated, showing LNM rates of 1.1% (low), 5.4% (intermediate), and 13.3% (high) in surgical cohorts [4,22].

In Korean practice, the decision for additional surgery after noncurative ESD is based on a comprehensive evaluation of pathological risk factors (depth of invasion, lymphovascular invasion, and margin status) and patient factors (age, comorbidities, and patient preference) through a multidisciplinary discussion rather than strict adherence to the eCura score [1]. European guidelines similarly emphasize multidisciplinary team discussions for high-risk resections [3]. The eCura system serves as a useful tool for counseling patients and guiding clinical decision-making but should not be interpreted as the sole determinant of management.

Globally, ESD has been established as the standard treatment for early gastric neoplasms, achieving excellent long-term survival comparable to that of surgery. The indications for ESD have expanded beyond classical absolute criteria to include larger differentiated lesions and, with caution, selected undifferentiated-type lesions including signet ring cell carcinoma in highly selected cases (≤2 cm, mucosal, no ulceration). However, there exist significant differences among international guidelines regarding terminology and risk stratification, particularly for undifferentiated histology and minimal submucosal invasion.

While technical difficulties remain, the use of various traction methods (clip-with-line, double-clip, pocket-creation) and appropriate accessories can enhance success rates, although comparative effectiveness data remain limited, and method selection should be individualized.

Clinicians must be vigilant of adverse events, including immediate and delayed complications. Recent updates in antithrombotic management emphasize drug-specific DOAC withdrawal periods based on renal function, continuation of aspirin for most cases, and a strong recommendation against heparin bridging for DOACs [17,18].

Finally, a comprehensive post-resection surveillance based on pathological curability is essential. For noncurative resections, risk stratification tools such as the eCura system provide valuable guidance, but management decisions should be made through multidisciplinary discussions considering patient-specific factors, particularly in elderly or high-risk surgical candidates [1,3,4].