INTRODUCTION
Esophagogastroduodenoscopy (EGD) is commonly used to evaluate gastrointestinal (GI) symptoms and to screen for cancer. During EGD, subepithelial lesions (SELs), previously known as submucosal tumors, are often encountered incidentally by endoscopists. SEL originate as a lesion, including tumor, bulge, or impression from the area underneath the epithelium of the GI tract, including the muscularis mucosae, submucosa, and muscularis propria. SEL can range from benign lesions, such as lipomas, leiomyomas, and duplication cysts, to malignant tumors, including gastrointestinal stromal tumors (GISTs) and neuroendocrine tumors.
SELs are found in 1 in every 300 endoscopies [
1–
3], with a higher frequency in Korea and Japan, where EGD is included in cancer screening [
4]. The stomach is the most commonly involved organ, followed by the esophagus [
5,
6], whereas the duodenum is the least involved organ in the upper GI tract [
5,
7].
Recent advancements in endoscopic procedures have rendered endoscopic resection (ER) a potential treatment option for upper GI SELs [
8–
10]. However, ER for duodenal lesions is challenging because of the anatomical characteristics of the duodenum, including a relatively thin muscle layer and narrow luminal diameter, which increases the risk of perforation and makes endoscopic maneuverability difficult. Surgical resection of duodenal SELs is also technically challenging because of their close proximity to major systemic and splanchnic vessels, the pancreas, and biliary organs, as well as their retroperitoneal location. Therefore, careful consideration is necessary when selecting duodenal SELs for endoscopic or surgical resections.
While several management plans and guidelines have been proposed for gastric SELs, limited data and information are available for the management of duodenal SELs, owing to a lack of knowledge about their natural course. This study aimed to evaluate the natural course of duodenal SELs and suggest appropriate management strategies.
DISCUSSION
Duodenal SELs are usually discovered incidentally during routine EGD and most are small (usually less than 2 cm) and asymptomatic. However, patients with SELs can present with GI bleeding, abdominal pain, or obstruction, for which resection is recommended [
2,
3,
5]. Nevertheless, there are limited data and knowledge about the natural course of asymptomatic duodenal SELs. The European Society of Gastrointestinal Endoscopy stated that there is insufficient evidence to guide recommendations for duodenal SELs and suggested that obtaining a definitive diagnosis is necessary for making further decisions [
5]. However, it is difficult to diagnose all asymptomatic duodenal SELs identified during routine EGD in a clinical setting.
In our study, duodenal SELs were identified in 0.39% of the cases (1,713/443,533). Among 396 asymptomatic duodenal SELs with a minimal follow-up period of 6 months, the majority (96.0%) did not exhibit substantial changes in size, morphology, or echogenicity during a median follow-up duration of 72.5 months (IQR, 37.7–111.3 mo). On the other hand, 16 (4.0%) SELs showed substantial changes at a median period of 35.1 months (IQR, 21.7–51.4 mo): 14 showed size increments and 2 showed surface ulceration. Among these cases, pathological confirmation was obtained for five SELs using EUS-FNAB, ER, or surgery, and only two (0.5%) were diagnosed as GISTs.
Several studies investigated the natural course of upper GI SELs. Gill et al. [
11] found that 86.3% (44/51) of upper GI SELs, including two duodenal SELs, demonstrated changes in size and/or echogenicity at a mean period of 29.7 months, whereas none of the duodenal SELs showed changes. Bruno et al. [
12] reported that 89.4% (42/47) of upper GI SELs, including four duodenal SELs, did not show changes in size and echogenicity. Lim et al. [
7] observed that 96.8% (244/252) of upper GI SELs, including 18 duodenal lesions, showed no changes over a mean duration of 59.1 months. Similarly, Song et al. [
13] reported that 96.4% (920/954) of upper GI SELs showed no interval changes during a median period of 47.3 months, and 96.2% (126/131) of duodenal SELs did not exhibit size increments during follow-up, which is consistent with the findings of our study. Kim et al. [
14] reported that 85.6% (575/672) of the upper GI SELs did not show changes at a mean period of 68 months, and 90.0% (81/90) of the duodenal SELs did not show any changes. Unfortunately, these studies predominantly concentrated on gastric SELs and had a limited scope in addressing the natural course of duodenal SELs.
EUS is a useful tool for the differential diagnosis of SELs by evaluating the layer of origin, size, and echogenic characteristics [
3,
15,
16]. EUS has demonstrated a sensitivity of 92% for distinguishing SELs from extrinsic compressions [
16]. However, differentiating between benign and malignant SELs using EUS alone poses challenges due to its relatively lower sensitivity (64%) and specificity (80%) [
2,
17]. Furthermore, the interpretation of EUS findings can vary between operators, and there may be poor interobserver agreement, particularly when assessing echogenic features suggestive of malignancy (echogenic foci, cystic spaces, irregular borders, or heterogeneity).
Therefore, pathological confirmation is often necessary for a definite diagnosis of SELs [
3,
4]. Because SELs are located beneath the epithelium, mucosal biopsies using standard biopsy forceps often fail to acquire tissues [
18]. EUS can facilitate tissue acquisition through EUS-FNAB, which is a widely used method for sampling lesions in the GI tract [
3,
19]. EUS-FNAB allows for the sampling of all GI tract lesions and has reported accuracy rates of 80–90% [
20–
25]. However, EUS-FNAB for duodenal SELs can be challenging because of difficulties in maneuvering an angulated scope-tip position in the duodenum [
2]. Additionally, the diagnostic yield is poor for duodenal SELs compared to that for gastric lesions [
26,
27]. Thus, we believe that it is important to predict the malignant potential of SEL on the basis of endoscopic and EUS findings. In our study, an initial size ≥ 20 mm was a significant factor associated with substantial changes in SELs (
p = 0.016), which aligns with previous reports and guidelines.
To date, our study is the largest to focus solely on duodenal SELs rather than on gastric lesions, as previous studies have predominantly concentrated on gastric SELs. In addition, factors such as age, sex, location, and the layer of origin were not statistically significant in relation to substantial changes in SELs (
Table 2), whereas an initial size ≥ 20 mm was a significant factor. Of the 379 duodenal SELs, only two (0.5%) were identified as GISTs. This finding suggests that regular endoscopic follow-ups may be an appropriate strategy for the management of duodenal SELs.
Neuroendocrine tumors are a rare type of neuroendocrine cancer that occur less frequently in the duodenum than in other parts of the digestive tract. Duodenal neuroendocrine tumors are usually small [
28], with mean sizes ranging from 7 to 15 mm [
29]. Due to their malignant potential, several guidelines advocate resection, regardless of their small size [
5,
30]. Therefore, even when encountering small duodenal SELs (< 20 mm), careful examination is important to exclude the possibility of neuroendocrine tumors. Neuroendocrine tumors are typically characterized by rounded lesions with a yellowish or reddish color that differ from the surrounding mucosa and are generally diagnosed using endoscopic mucosal forceps biopsy [
5].
Our study had several limitations. First, this was a single-center retrospective observational study, which might have limited the generalizability of our results. Second, multiple endoscopists performed EGD and EUS to evaluate SELs. To minimize inter-observer variations, the EGD and EUS images were reviewed by a single investigator. Third, EUS, which can provide valuable information for SEL assessment, was not performed in all cases. Fourth, not all 16 SELs with substantial changes underwent pathological confirmation; tissue acquisition was attempted in only five cases, four of which were diagnosed pathologically as two GISTs and two lipomas. Some patients refused further evaluation or were lost to follow-up. In other cases, technical difficulties were encountered owing to the acute angulation position of EUS or small lesion size. Finally, owing to the retrospective design of this study, certain endoscopic characteristics of SELs, including consistency and mobility, could not be evaluated using EGD images, and therefore, were not available.
In conclusion, the majority of duodenal SELs did not exhibit substantial interval changes during long-term follow-up. Nevertheless, regular endoscopic follow-up is recommended for cases with an initial size of 20 mm or larger, considering a possibility of malignancy.