Scope
The ACG subepithelial-lesion guideline distinguishes incidental benign findings from lesions needing tissue diagnosis, surveillance or resection. Endoscopic ultrasound is central to assessing nonlipomatous lesions, but diagnosis often requires sampling. Most treatment recommendations are conditional and supported by very-low-quality evidence, so lesion site, layer, symptoms, pathology, procedural expertise and patient preferences remain important.
Written by House with AI assistance · AI-assisted source check 29 October 2026.
This is House’s summary of the cited guidance. The source check is not an independent clinician review. Open the publication for complete recommendations, exceptions, evidence grades and later amendments.
Use EUS and tissue diagnosis appropriately
- Prefer endoscopic ultrasound (EUS) over ordinary endoscopy or contrast-enhanced cross-sectional imaging for diagnosing nonlipomatous subepithelial lesions. This is a conditional recommendation with very-low-quality evidence, not a claim that every lesion requires the same investigation. [1]
- Combine EUS with tissue acquisition to improve diagnostic accuracy for solid nonlipomatous lesions; EUS appearance alone is not sufficient for reliable classification. The recommendation is conditional, with very-low-quality evidence. [1]
- The guideline does not prefer forward-viewing over oblique-viewing echoendoscopes for evaluating subepithelial lesions: this is a strong recommendation based on low-quality evidence. Instrument choice remains at the operator’s discretion. [1]
- Do not use bite-on-bite biopsy before EUS for routine lesion evaluation; this is a conditional recommendation with very-low-quality evidence. The source’s potential fallback use after EUS requires exclusion of vascular lesions and inability to perform the preferred needle-sampling approach. [1]
Choose an appropriate sampling approach
- For solid nonlipomatous lesions, favour EUS fine-needle biopsy alone or fine-needle aspiration with rapid on-site evaluation over aspiration without on-site evaluation. This is a conditional recommendation based on low-quality evidence. [1]
- If a definitive diagnosis before resection is necessary but EUS aspiration or biopsy is nondiagnostic, consider an unroofing technique. This conditional, low-quality recommendation concerns diagnostic failure, not a universal initial procedure. [1]
- There is no fixed minimum size below which needle sampling cannot be attempted. Manage small lesions individually using their location, accessibility, clinical history, surveillance risks and benefits, and whether primary endoscopic resection is feasible. This is an ungraded key concept. [1]
- When EUS is unavailable, contrast-enhanced imaging may add useful diagnostic information, but consider referral to an EUS-capable centre according to the differential diagnosis. This is an ungraded key concept rather than evidence that CT replaces EUS routinely. [1]
Distinguish observation from indications for resection
- Symptoms or gastrointestinal bleeding may justify resection regardless of lesion size without a prior tissue diagnosis. A large GIST is an exception: tissue confirmation may enable neoadjuvant imatinib to reduce tumour size before resection. This is an ungraded key concept. [1]
- If a lesion remains without tissue diagnosis or resection, establish surveillance unless there is high confidence that it has no malignant potential. Asymptomatic benign lesions do not require resection, and the guideline cannot specify definitive surveillance intervals. [1]
- When a muscularis-propria lesion of the oesophagus or gastro-oesophageal junction needs resection, consider either submucosal tunnelling endoscopic resection or surgery. The recommendation is conditional with very-low-quality evidence; it does not make resection mandatory for every such lesion. [1]
Keep gastric and nongastric GIST decisions distinct
- For gastric GIST smaller than 2 cm, evidence is insufficient to choose surveillance over resection. The guideline conditionally suggests resection for gastric GIST larger than 2 cm and for all nongastric GIST because of malignant potential; evidence quality is very low. [1]
- If a decision has been made to resect a GIST below 2 cm, endoscopic approaches may be acceptable alternatives to surgery. No particular endoscopic technique is established as superior. This is an ungraded key concept and does not resolve whether the lesion should first be resected. [1]
Match resection technique to NET subtype
- For type 1 gastric neuroendocrine tumours requiring resection, consider endoscopic mucosal resection (EMR) or endoscopic submucosal dissection (ESD). The recommendation is conditional with very-low-quality evidence. [1]
- For small, low-grade type 3 gastric NETs without radiological or EUS lymphadenopathy that are not treated surgically, favour ESD over EMR conditionally, with very-low-quality evidence. The accompanying discussion describes superficial lesions below 1 cm for this endoscopic approach. [1]
- For small, low-grade rectal NETs below 1 cm, neither EMR nor ESD is preferred by the formal recommendation. This is conditional with very-low-quality evidence; it should not be extended to larger or higher-grade tumours. [1]
Limitations and use
- The guideline leaves the surveillance-versus-resection choice for gastric GIST below 2 cm unresolved, and it does not establish a definitive surveillance-interval schedule. The exact 2 cm boundary is not assigned by its less-than/greater-than statements.
- This summary does not provide staging, systemic oncology regimens, complete NET subtype management or technical procedural instructions. Local EUS, pathology and advanced-endoscopy expertise may limit implementation in India.
This summary is selective and does not reproduce every recommendation. Apply the source’s population, setting and eligibility criteria, together with clinical judgement and local policy.
Summary sources
- Diagnosis and Management of Gastrointestinal Subepithelial Lesions
Selected recommendation material consulted
Publication details
- Issuing organisation
- ACG
- Resource type
- Clinical guideline · House clinical summary
The source check confirms publication details and the linked edition. It does not certify every recommendation or its suitability for an individual patient. Publisher access conditions may apply.