EUS-CNB Versus EUS-SINK for Diagnosis of Upper Gastrointestinal (UGI) Subepithelial Tumors

Upper gastrointestinal (GI) subepithelial tumors (SETs) are tumors arising from subepithelial layers of esophageal, gastric or duodenal wall, mostly from the submucosa and muscular layer. The incidence of SETs on routine endoscopy is 0.36% The differential diagnosis of SETs include, though are not limited to: lipoma, leiomyoma, aberrant pancreas, varices, carcinoid, gastrointestinal stromal tumors (GISTs), and lymphomas . Therefore, a correct diagnosis of these tumors is important to guide subsequent management. These lesions are often not accurately diagnosed on cross-sectional imaging . Endoscopic ultrasound (EUS) aids in narrowing the differential diagnosis of the lesion as it is often able to establish the layer of origin . However, an accurate diagnosis and targeted therapy is not made solely on the morphological features but on histologic type and at times mitotic index. Thus the need for techniques to obtain histology is beneficial in guiding management. Since standard endoscopy with pinch biopsies of the overlying mucosa often fails to provide an adequate sample for analysis, multiple other modalities to sample the lesion have been utilized: EUS-guided fine needle aspiration (EUS-FNA), EUS-guided core needle biopsy (EUS-CNB), bite-on-bite forceps biopsies, EUS-guided single-incision with needle knife (SINK) and endoscopic resection. EUS-FNA is now considered to be the usual method of sampling; however, the diagnostic yield is low: 38% to 82% . Moreover, EUS-FNA often provides insufficient specimens which may not allow for immunohistochemistry that is often essential for diagnosis . Thus EUS-CNB has been assessed for the purpose of obtaining a core sample which allows for histological assessment. Published data reveals a diagnostic (though not histologic) yield using EUS-CNB of 75% In 2011, the SINK technique for sampling was presented with a reported diagnostic accuracy of 92.8% \[8\]. The technique utilizes a conventional needle-knife connected to an electrosurgical unit. A 6 to 12-mm mucosal incision is made over the lesion. Then conventional biopsy forceps are introduced to obtain 3-5 samples. Subsequently, the incision is closed with 2 to 3 endoclips. The purpose of this study is to prospectively compare the efficacy and safety of EUS-CNB with SINK in patients with upper GI SETs. The investigators hypothesis is that the SINK technique will be superior to the EUS-CNB in obtaining a histological specimen. The results of the study would provide data which may improve the diagnostic ability for SETs. This in turn will guide appropriate surveillance or management (surgical or endoscopic) for patients with these lesions.