Feasibility and clinical value of linear endoscopic ultrasonography imaging in the lower gastrointestinal subepithelial lesions.

This was a retrospective study. From August 2019 to April 2023, a total of 69 patients with lower gastrointestinal subepithelial lesions who underwent linear EUS were collected from the Endoscopy Center of the Second Affiliated Hospital of Anhui Medical University. Before undergoing linear EUS, all patients had previously undergone colonoscopy. Excluding 7 patients with unknown diagnosis (who refuse further examination), 3 patients with no abnormal EUS findings, and 3 patients failed the procedure, 56 patients were included in the final diagnostic performance analysis. Linear EUS (EU-ME2, GF-UCT260, Olympus, Tokyo, Japan; EG-3270UTK, Pentax, Tokyo, Japan) was utilized for the examination. The probe frequency is dynamically adjusted depending on the distance of the lesion from the probe and whether there is extraluminal compression, most of which is 7.5 MHz, occasionally 5 MHz or 9 MHz. All patients signed informed consent for EUS examination. To prepare for linear EUS, 654–2 (10 mg) or diazepam (10 mg) was administered as premedication 30 min before the procedure. Given that the accuracy of EUS diagnosis heavily relies on the experience of the endoscopists, all patients in this study underwent examination and diagnosis by two experienced endoscopists (each with experience of > 1000 EUS examinations) to minimize bias. Clinical and colonoscopic information was given before EUS evaluation. However, endoscopists were blinded to the pathological results of these lesions. The procedure was conducted through an endoscope inserted into the anus. A downward angle allowed for viewing the rectal lumen, followed by insufflation with air and advancing the endoscope. The transition from the rectum to the recto-sigmoid junction was relatively straightforward. However, upon reaching the sigmoid colon, splenic flexure, and hepatic flexure, the direction of endoscopic advancement may not have been visible due to relatively large angles, which could be mitigated by adjusting body position and applying abdominal pressure. It was crucial to navigate these areas slowly and gently. Rotating the endoscope axis while entering the mirror was essential. In the descending colon, transverse colon, and ascending colon, visualizing the intestinal lumen before proceeding forward was necessary. Once the lesion was reached and the gas was evacuated, the water pump was employed to inject saline into the intestinal cavity, completely immersing the lesion. Throughout the process, it was important to bear in mind the distinction between the oblique endoscope view and direct vision. The data, encompassing basic clinical information, linear EUS features, technical success rate (the success rate of operating linear EUS to reach the lesion location), were retrospectively collected and analyzed. The diagnosis and management were determined according to guidelines of the American Society for Gastrointestinal Endoscopy (ASGE) 3 . Subsequent treatments were chosen based on linear EUS imaging diagnosis of lesion, including options such as follow-up, biopsy, EUS-FNA, endoscopic resection, or surgical excision. Biopsy and postoperative pathological results were compared with the diagnoses obtained via linear EUS imaging. Patients not undergoing immediate intervention were periodically followed up. The primary outcome of this study was the diagnostic accuracy of linear EUS imaging in lower gastrointestinal subepithelial lesions. The secondary outcome included the assessment of technical success rate, safety and feasibility. The data processing was performed using the SPSS 22.0 statistical software. Categorical data are displayed as number (n) or percentage (%) and were compared using the Chi-Square Test . Quantitative data are expressed as the mean ± standard deviation (SD) and were compared using the t -test. A p value less than 0.05 was considered statistically significant. This study was reviewed and approved by the Institutional Review Board of the Second Affiliated Hospital of Anhui Medical University (SL-YX2023-181), and conforms to the principles stated in the Declaration of Helsinki.

A total of 69 patients with lower gastrointestinal subepithelial lesions underwent examination by linear EUS, resulting in 3 procedure failures. Exception the rectum, the technical success rate of linear EUS in reaching the lesion location was 90.6% (29/32). The final diagnostic performance analysis comprised 56 patients, including 29 males and 27 females, with an average age of 53.0 ± 12.04 years (range: 22–79 years). The average diameter of all lesions measured 15.95 ± 9.82 mm (range: 4.6–45.1 mm). The most frequent locations of the lesions were the rectum (37/56, 66.1%) and sigmoid colon (7/56, 12.5%), followed by transverse colon (4/56, 7.1%) and ascending colon (4/56, 7.1%) (Table 1 ). Table 1 Distributions of subepithelial lesions in the lower gastrointestinal tract. Lesions Cecum Ascending colon Hepatic flexure Transverse colon Descending colon Sigmoid colon Rectum Total Percentage NET – – – – – – 12 12 20.3 Lipoma 1 4 – 1 – 1 1 8 13.6 Extraluminal compression – – 1 – 1 1 5 8 13.6 Cancer – – – – – 1 6 7 11.9 Polyps – – – – – 2 2 4 6.7 Inflammation – – – – – – 4 4 6.7 Cyst – – – 3 – – – 3 5.1 Varix – – – – – – 2 2 3.4 Glomus – – – – – 2 – 2 3.4 PCI – – 1 – – – 1 2 3.4 GIST – – – – – – 2 2 3.4 Schwannoma – – – – – – 1 1 1.7 Endometriosis – – – – – – 1 1 1.7 Total 1 4 2 4 1 7 37 56 – Percentage 1.8 7.1 3.6 7.1 1.8 12.5 66.1 – 100 Distributions of subepithelial lesions in the lower gastrointestinal tract. Based on the findings from endoscopy and pathological results, the most prevalent types of lower gastrointestinal subepithelial lesions were neuroendocrine tumor (NET) (12/56, 20.3%), all of which were situated in the rectum. Subsequently, other common types included lipoma (8/56, 13.6%), extraluminal compression (8/56, 13.6%), and cancer (7/56, 11.9%). Additionally, various other observed lesions in the study comprised polyps, inflammation, cysts, varices, glomus, among others. These outcomes are visually displayed in Fig.  1 and Table 1 . Figure 1 Endoscopic and ultrasonographic images of lower gastrointestinal subepithelial lesions. ( A ) Neuroendocrine tumor (NET); ( B ) Lipoma; ( C ) Renal cyst extraluminal compression; ( D ) Cancer; ( E ) Inflammation; ( F ): Cyst; ( G ) Varix; ( H ): Glomus; ( I ) Pneumatosis cystoides intestinalis (PCI); ( J ) Gastrointestinal stromal tumor (GIST); ( K ) Schwannoma; ( L ) Endometriosis. Endoscopic and ultrasonographic images of lower gastrointestinal subepithelial lesions. ( A ) Neuroendocrine tumor (NET); ( B ) Lipoma; ( C ) Renal cyst extraluminal compression; ( D ) Cancer; ( E ) Inflammation; ( F ): Cyst; ( G ) Varix; ( H ): Glomus; ( I ) Pneumatosis cystoides intestinalis (PCI); ( J ) Gastrointestinal stromal tumor (GIST); ( K ) Schwannoma; ( L ) Endometriosis. Linear EUS revealed extraluminal compression in 8 patients, attributed to external pressure organs include renal cysts, uterus and prostate. Among the remaining 48 patients, the linear EUS features varied. Most lesions ranged in diameter from 1 to 3 cm ( χ 2  = 18.750, p  < 0.001) (Table 2 ). Out of 12 cases of NET, 9 were less than 1 cm in diameter ( χ 2  = 15.750, p  < 0.001). Furthermore, lipomas were concentrated in the 1 to 3 cm in diameter range ( χ 2  = 10.500, p  = 0.005). As depicted in Table 3 , the most common sites of origin of NET were the muscularis mucosa and submucosal with a distinct border ( χ 2  = 16.889, p  = 0.001). Nearly all lipomas originated from the submucosa with a distinct border ( χ 2  = 22.667, p  < 0.001). Additionally, all cancer and polyps originated from the mucosa, with disrupted mucosal continuity. Cysts, varices, pneumatosis cystoides intestinalis (PCI) and schwannoma originated from the submucosa, while glomus, gastrointestinal stromal tumor (GIST), and endometriosis originated from the muscularis propria. In mucosal lesions, polyps and inflammation appeared hyperechoic while cancers were hypoechoic. The majority of NETs were visualized as hypoechoic, with 2 cases being hyperechoic ( χ 2  = 30.222, p  < 0.001). All lipomas were hyperechoic ( χ 2  = 32.000, p  < 0.001). Conversely, all cysts were anechoic. The comprehensive results can be found in Table 4 . Table 2 Sizes of subepithelial lesions in the lower gastrointestinal tract. Lesions   3 cm Total P X 2 NET 9 3 – 12  < 0.001 15.750 Lipoma 2 6 – 8 0.005 10.500 Cancer – 4 3 7 0.062 5.571 Polyps 1 2 1 4 0.687 0.750 Inflammation 3 1 – 4 0.072 5.250 Cyst – 3 – 3 / / Varix – 2 – 2 / / Glomus – 2 – 2 / / PCI – 2 – 2 / / GIST – – 2 2 / / Schwannoma 1 – – 1 / / Endometriosis – 1 – 1 / / Total 16 26 6 48  < 0.001 18.750 Table 3 Originating layers of the lesions on endoscopic ultrasonography (EUS). Lesions Mucosa layer Muscularis mucosa layer Submucosal layer Muscularis propria layer Total P X 2 NET – 5 7 – 12 0.001 16.889 Lipoma – 1 7 – 8  < 0.001 22.667 Cancer 7 – – – 7  < 0.001 28.000 Polyps 4 – – – 4 / / Inflammation 2 – 2 – 4 / / Cyst – – 3 – 3 / / Varix – – 2 – 2 / / Glomus – – – 2 2 / / PCI – – 2 – 2 / / GIST – – – 2 2 / / Schwannoma – – 1 – 1 / / Endometriosis – – – 1 1 / / Total 13 6 24 5 48  < 0.001 25.556 Table 4 Echogenicity of the lesions on endoscopic ultrasonography (EUS). Lesions Hyperecho Isoecho Hypoecho Anecho Total P X 2 NET 2 – 10 – 12  < 0.001 30.222 Lipoma 8 – – – 8  < 0.001 32.000 Cancer – 1 6 – 7  < 0.001 18.857 Polyps 4 – – – 4 / / Inflammation 3 1 – – 4 / / Cyst – – – 3 3 / / Varix 1 – – 1 2 / / Glomus – 1 1 – 2 / / PCI – – 2 – 2 / / GIST – – 2 – 2 / / Schwannoma – – 1 – 1 / / Endometriosis – – 1 – 1 / / Total 18 3 23 4 48  < 0.001 33.556 Sizes of subepithelial lesions in the lower gastrointestinal tract. Originating layers of the lesions on endoscopic ultrasonography (EUS). Echogenicity of the lesions on endoscopic ultrasonography (EUS). Out of these, 36 patients underwent histological confirmation by EUS-FNA (3/36), biopsy (5/36), endoscopic resection (25/36), or surgical excision (3/36) pathology. The diagnostic accuracies of linear EUS for lower gastrointestinal lesions are presented in Table 5 . Postoperative pathological results and biopsies from 29 patients (29/36) were entirely consistent with the initial linear EUS diagnosis, demonstrating an 80.6% diagnostic accuracy. Linear EUS exhibited a 100% diagnostic accuracy for lipoma (1/1), cyst (2/2), glomus (1/1), and GIST (2/2), and achieved a 91.7% accuracy for NET (11/12). Additionally, the diagnostic accuracy of linear EUS was 85.7, 75 and 75% for cancer (6/7), polyps (3/4) and inflammation (3/4), respectively. There were 7 cases of misdiagnosis, which included 1 case each of inflammation, PCI and schwannoma misdiagnosed as NET, 1 case of NET and polyps misdiagnosed as lipoma, 1 case of cancer misdiagnosed as polyps, and 1 case of endometriosis misdiagnosed as GIST. Table 5 Diagnostic rates of endoscopic ultrasonography (EUS) diagnosis based on final histopathology of the lesions. Pathological results Number of pathological diagnoses Number of EUS diagnoses Number of consistent diagnoses between pathology and EUS Diagnostic consistency rate (%) NET 12 14 11 91.7 Lipoma 1 3 1 100 Cancer 7 6 6 85.7 Polyps 4 4 3 75 Inflammation 4 3 3 75 Cyst 2 2 2 100 Glomus 1 1 1 100 PCI 1 – – 0 GIST 2 3 2 100 Schwannoma 1 – – 0 Endometriosis 1 – – 0 Total 36 36 29 80.6 Diagnostic rates of endoscopic ultrasonography (EUS) diagnosis based on final histopathology of the lesions. Among these were 20 cases involving submucosal lesions that did not undergo endoscopic therapy or surgery, excluding cases with extraluminal compression. Out of these, 12 cases were monitored through endoscope or EUS for a span of 6–36 months. The findings revealed that the lesions reminded unchanged during the follow-up period. Notably, no complications associated with lower gastrointestinal linear EUS examinations were reported among any of the patients.

In summary, this retrospective study investigated the effect of linear EUS imaging in lower gastrointestinal subepithelial lesion. Our results demonstrated linear EUS exhibits high diagnostic accuracy for subepithelial lesions in the lower gastrointestinal tract. The evidence suggests that linear EUS is technically feasible and safe in the lower gastrointestinal tract. It can effectively aid in the differentiation of such lesions. Given its significant clinical application value, promotion of the use of linear EUS is warranted.

With the advancement of colonoscope and imaging examination, the reported incidence of subepithelial lesions in the lower gastrointestinal tract has increased 4 , 13 . In instances where lower gastrointestinal subepithelial lesions are encountered during colonoscopy and prove difficult to identify, EUS can be considered an ideal method for further examination as it furnishes valuable information concerning the location of the lesions (intramural or extramural), size, and echogenic characteristics 14 . There are three EUS methods for detecting lower gastrointestinal subepithelial lesions: miniprobe EUS, radial EUS and linear EUS 15 . Compared with the other two types, the great advantage of linear EUS is that lesions and deeper layers can be better visualized on the same image. An additional advantage of linear EUS is the ability to sample the tissue by EUS-FNA/FNB 8 . Currently, most studies have concentrated on mini-probe EUS or radial EUS 4 , 11 – 13 , 16 , with fewer investigations employing linear EUS to explore lower gastrointestinal lesions. In a recent study, linear EUS was solely utilized in the rectum 11 . However, there are minimal relevant studies on the application of linear EUS for the entire lower gastrointestinal tract due to equipment requirement and operational challenges. In our study, linear EUS was employed to investigate lower gastrointestinal subepithelial lesions. The findings revealed that out of 69 patients who underwent linear EUS, 3 experienced procedural failures. Challenges were encountered in the hepatic region (1 patient) and sigmoid colon (2 patients), respectively. Except for the rectum, the operational success rate of linear EUS stood at 90.6% (29/32). This high technical success rate suggests that linear EUS is a viable method for examining the lower gastrointestinal tract, with the exception of the rectum. The study demonstrates that linear EUS is capable of clearly displaying the adjacent structure of the infrasonic and pelvic compartments. Imaging from the upper rectum allows visualization of the left infrasonic compartment, retroperitoneal structures, and bowel loops 17 . However, there are numerous subepithelial lesions in the lower gastrointestinal tract above the rectum, which also require identification using linear EUS. When a subepithelial lesion is detected, accurate determination of its etiology is crucial as each type of subepithelial lesion requires different treatments, follow up, and prognosis. If extraluminal compression arises from a normal adjacent structure or organ, no further examination is necessary. In our study, linear EUS revealed extraluminal compression in 8 patients, attributed to external pressure organs such as renal cysts, uterus, and prostate, for which no further treatment was required. Meanwhile, 3 patients exhibited no abnormal findings under linear EUS. On one hand, it is possible that the 3 lesions may have been missed due to the existence of blind areas in the oblique endoscope view. On the other hand, the potential for extraluminal compression could not be excluded. Based on the features observed with linear EUS, it is possible to diagnose cysts or lipomas without further tests or biopsies. Follow-up may only be necessary for selected patients, without invasive or aggressive intervention. However, when a hypoechoic subepithelial lesion is visible in the submucosal or muscularis propria, the differential diagnosis may include NET, GIST, leiomyoma, endometriosis, or schwannoma. Prior studies have shown that the diagnostic accuracy for lesions in the submucosal or muscularis propria layer using EUS alone is only around 50% 18 . In our study, pathological diagnoses were compared with the results obtained using linear EUS. The pathological outcomes of 29 patients were entirely consistent with the linear EUS diagnosis, resulting in an 80.6% (29/36) diagnostic accuracy. Rare lesions such as glomus, schwannoma, and endometriosis lacked distinctive features in linear EUS and appeared similar to common submucosal lesions, leading to potential misdiagnoses under linear EUS. Furthermore, the diagnostic accuracy of linear EUS is closely linked to operator skill, equipment, and clinical experience 19 . Thus, the diagnosis of subepithelial lesions must be combined with the patient’s medical history and relevant tests, especially pathological examinations 20 . Consequently, linear EUS-guided tissue acquisition may play a crucial role in distinguishing between these subepithelial lesions. The oblique endoscope view of linear EUS can lead to increased entry difficulty and a heightened risk of perforation. In the lower gastrointestinal tract, negotiating the excessively long and freely mobile sigmoid colon is challenging, and the acute angles at the hepatic flexure and splenic flexure further exacerbate passage difficulties. During the operation, operators must delicately manipulate and control the lens while avoiding large movements, abrupt rotations, and forceful entries, especially through the aforementioned areas. The most critical aspect is to refrain from blindly conducting forced operations. Intravenous anesthesia or sedative drugs are recommended for patients undergoing linear EUS. In our study, all patients received antispasmodic or sedative treatment before the examination, and no significant discomfort was experienced during the procedure. Notably, no complications were reported among any of the patients who underwent lower gastrointestinal linear EUS examinations. Taken together, these results provided evidence that linear EUS is technically feasible, safe, and exhibits high diagnostic accuracy for subepithelial lesions in the lower gastrointestinal tract. This study presents several limitations. Firstly, it was a retrospective study, potentially introducing selection bias during data collection. Secondly, our study was confined to a single-center. Thirdly, the small sample size further limited the scope of the study. Future analyses would benefit from larger prospective, multicenter studies to provide additional insights and comprehensive analysis of the findings.