AGA Offers Key Guidance on Managing Subepithelial GI Lesions

David A. Johnson, MD


January 10, 2023

This transcript has been edited for clarity.

Hello. I'm Dr David Johnson, professor of medicine and chief of gastroenterology at Eastern Virginia Medical School in Norfolk, Virginia.

Subepithelial lesions of the gastrointestinal (GI) tract are masses, bulges, or impressions in the GI lumen. When viewed endoscopically, they're covered with normal-appearing mucosa.

Subepithelial lesions are not uncommon, found in approximately 1 in 300 endoscopies, two thirds of which are in the stomach. They represent a heterogeneous composite, ranging from those with malignant potential to nonneoplastic lesions.

The majority are small and found incidentally, without any signs or symptoms. If there are signs or symptoms, they are most commonly related to bleeding or abdominal pain. But again, that is very unusual, and they rarely cause bowel obstruction or present with metastasis.

Therefore, for the majority of these lesions, our proposed management strategies really depend on their malignant potential.

Expert consensus recommendations on subepithelial lesions were issued recently from the American Gastroenterological Association (AGA) and provide us with some excellent guidance going forward regarding how to manage them clinically. I want to highlight a few key messages from these guidelines for you.

Diagnostic Endoscopy

The authors highlight several diagnostic methods, including endoscopy, forceps, biopsy, and endoscopic ultrasound. For each of these, they also provide key messages.

A lot of these lesions are seen in the stomach. The endoscopic appearance of pancreatic rests are very commonly central umbilication lesions, typically < 1 cm or approximately 6-10 mm in size. They're also located within 2-6 cm proximity of the pylorus and almost always on a greater curvature, which is very classic type of description. This may require a mucosal biopsy, but once that's defined, then you don't need to do anything further.

Sometimes, we see lesions in the fundus, which creates a question whether it's a submucosal lesion or an extrinsic compression (from adjacent organs, such as the spleen). Assessing this sometimes requires rotating the patient or placing them more in an upright position. The sensitivity of this maneuver is low, but it's something that the authors do recognize may be somewhat helpful.

Forceps Biopsy

Because these lesions are submucosal by nature, mucosal biopsies using a standard biopsy are almost never diagnostic. With this in mind, the authors recommend a couple of different options.

One is called a bite-on-bite "tunneling" biopsy, typically using large or jumbo forceps. During this technique, you typically use a three-on-three biopsy approach where you go in, biopsy, open it again, biopsy, push in a little further, biopsy, and push in a little further.

These biopsies have been described as having a diagnostic yield of 30%-40%. For lesions arising from the submucosa, or third layer, the yield is comparable to endoscopic ultrasound.

The advantage is lower when lesions are deeper and arising from the muscularis propria. There are a variety of techniques discussed in these guidelines, including unroofing techniques using needle knife, snare, or band ligation.

But again, the three-on-three tunneling biopsy technique is something that I found very helpful.

Endoscopic Ultrasound

The authors note that endoscopic ultrasound is the method of choice for characterizing submucosal lesions of the GI tract as an adjunct to the standard endoscopy.

Echogenicity is really a key aspect here. The echogenic features are typically anechoic lesions, usually represented as fluid-filled structures such as cysts, varices, or lymphangiomas. Hypoechoic lesions represent a variety of pathologic causes that include mesenchymal tumors, granular cell tumors, neuroendocrine tumors, metastatic disease, infiltrative diseases, and inflammation. Also, don't forget endometriosis when you're looking at the rectum. Hyperechoic lesions typically are lipomas or fibrolipomas. Mixed echogenicity may represent pancreatic rest, malignant tumors, or even GI tract wall abscesses.

The size of this lesion and the presence of vascular involvement are additional features that actually help predict the malignant potential of these subepithelial lesions.

The sensitivity and specificity of endoscopic ultrasound in predicting this malignant potential is in the range of 64% and 80%, respectively. The problem here is that it's operator dependent, making this a key area where artificial intelligence may allow for better sensitivity and specificity. But again, you need to have an expert to make sure that we're getting the correct and the most appropriate information.

Common Subepithelial Lesions

When it comes to differentiating among subepithelial lesions, in particular submucosal lesions like GI stromal tumors, the key discriminant is typically size. Size matters, particularly for lesions < 2 cm.

In the stomach, the gastric intestinal tumor < 2 cm can be surveyed using endoscopic ultrasound. The optimal surveillance interval is not well defined. Annual surveillance is the current recommendation. Larger lesions certainly can be considered for resection.

Leiomyomas are almost always benign and found in the esophagus. They represent about two thirds of the spindle cell tumors of the esophagus. Differentiation here requires histologic characterization. Again, this is a lesion type that's often found to be more asymptomatic, and they generally don't require surveillance. But if there is any question, these need to be surveyed using histologic sampling with fine-needle aspiration and fine-needle biopsy.

Carcinoids are the lesion type that the authors raise as requiring particular attention. Gastric carcinoids typically are divided into three types.

Type 1 is in the setting of chronic autoimmune gastritis and is associated with hypergastrinemia and high gastric pH.

Type 2 develops in the setting of gastrinomas, with hypergastrinemia and low gastric pH, and could be part of the multiple endocrine neoplasia-1 syndrome. They're rarely metastatic, with typically low mitotic indices and generally not greater than 1-2 cm.

Type 3 is the red flag. They're not derived from any underlying gastric pathology and not related to gastrin secretion. The problem with type 3 is that they often require radical resection, and a subset of smaller lesions may be followed. To learn more about type 3 carcinoids, I suggest you review these guidelines. When you start to make discriminant questions, make sure you have your histopathologist well attuned to what you're looking for.

Duodenal carcinoids are often nonfunctioning. These are localized lesions typically found coincidentally. Endoscopic resection is preferred, if possible (once again, performed by experts), especially for lesions ≤ 10 mm in size, limited to the submucosa. Larger lesions with more extensive involvement require resection performed at a high-volume center. They may also warrant surgical referral because the perforation or bleeding risk is inordinately high in the duodenal area.

Rectal carcinoids are something that I'm sure you've probably all encountered. The management of these depends on the size. The current recommendation is that completely resected lesions < 1 cm do not require further evaluation or surveillance. Lesions that are 1-2 cm are to be resected using endoscopic techniques or transanal surgery if their stage is at T1 lesion without nodal involvement. After resection, these lesions require surveillance with endoscopy, an endoscopic ultrasound, or MRI at the 6- to 12-month interval.

Granular cell tumors are most commonly seen in the esophagus, mainly in the submucosal layer. These are small lesions with a benign behavior, which can be surveyed by endoscopic evaluation alone or with endoscopic ultrasound. These typically do not require fine-needle aspiration; the same is true for duplication cysts, as there is a risk for mediastinitis. If duplication cysts have a classic appearance, they do not need further surveillance.

The AGA's Best Practice Recommendations

I'd like to leave you with the AGA's best practice recommendations.

First is to use bite-on-bite biopsies if you have histologic questions at the start.

They then go on to state that endoscopic ultrasound is the evaluation of choice for indeterminate subepithelial lesions in the GI tract and/or if nondiagnostic tissue found by the biopsy forceps.

The next recommendation is that, for lesions arising from the submucosa, sampled by tunnel biopsies, and for which you have a definition of the diagnosis, you don't need to further evaluate with fine-needle aspiration, fine-needle biopsy, or advanced endoscopic techniques, such as unroofing or endoscopic submucosal resection. Again, these more advanced procedures must be done in the hands of experts at high-volume centers.

The next piece of advice is that the management these lesions depends on their size, histopathology, malignant potential, and presence of symptoms. The bottom line is that if there's signs or symptoms, bleeding or pain, that heightens the need for resection.

You also need to consider the lesion's endoscopic appearance. For example, with a typical lipoma, you'll see the "pillow sign" when you push with the forceps. There are also specific appearance considerations with pancreatic rests and normal mucosal biopsies, which do not need further evaluation or surveillance. Duplication cysts would be in the same category.

For lesions rising from the muscularis propria that are < 2 cm in size, the recommendation is still to survey these with endoscopic ultrasound. Although it's not based on high-quality evidence, the current standard recommendation is to do this at 1 year.

The next recommendation is that gastrointestinal stromal tumors > 2 cm should be considered for resection. Again, I'd ask you to review these guidelines, particularly as it relates to carcinoids and the discriminants to apply based on whether they're type 1, 2, or 3.

Finally, I'd like to reiterate that the subepithelial lesions that are ulcerated, bleeding, or causing symptoms should be considered for resection, once again in the hands of an expert.

In conclusion, I'd bookmark these AGA expert consensus recommendations and refer to it the next time you have a subepithelial lesion in the GI tract. This is the best practice evidence we have to date.

Hopefully, you've found these comments helpful. See you next time.

David A. Johnson, MD, a regular contributor to Medscape, is professor of medicine and chief of gastroenterology at Eastern Virginia Medical School in Norfolk, Virginia, and a past president of the American College of Gastroenterology. His primary focus is the clinical practice of gastroenterology. He has published extensively in the internal medicine/gastroenterology literature, with principal research interests in esophageal and colon disease, and more recently in sleep and microbiome effects on gastrointestinal health and disease.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.