Sub epithelial lesions (SELs) of the gastrointestinal (GI) tract are commonly identified during routine endoscopy. Most of these lesions are benign. However because there is the potential for malignant transformation it is important to correctly identify the lesion in order to determine if any further therapy and/or surveillance is necessary for the patient, particularly for gastrointestinal stromal tumors (GISTs).

Obtaining a definitive diagnosis for SELs is often difficult since biopsies of the normal overlying surface mucosal layer are typically normal. EUS-FNA is the standard method by which a biopsy-proven diagnosis is obtained for most SEL's. However, the yield for a definite diagnosis from EUS-FNA for SELs is often suboptimal. Recently a new biopsy method, called "single incision needle-knife" (SINK) was introduced that may prove more useful in determining a definitive diagnosis. Furthermore, recent advances in core biopsy needles for EUS offer the hope for improved outcomes with EUS-guided fine-needle biopsy (FNB). However, it remains unclear whether superior diagnostic outcomes are obtained using the new SINK biopsy method or using new EUS-FNB core needles.

Subepithelial lesions of the gastrointestinal (GI) tract are commonly identified during routine endoscopy. Most of these lesions are benign. However because there is the potential for malignant transformation it is important to correctly identify the lesion in order to determine if any further therapy and/or surveillance is necessary for the patient.

Leiomyoma, neural origin tumors, lipoma, duplication cyst, pancreatic rest, inflammatory fibroid polyp, granular cell tumor are considered benign SELs. Gastrointestinal stromal tumor, lymphoma, carcinoid, metastatic carcinoma, glomus tumor are malignant or potentially malignant lesions. Obtaining a diagnosis for SELs is often difficult since biopsies of the normal overlying surface mucosal layer are typically normal. Jumbo forceps biopsy, bite-on-bite technique with conventional biopsy forceps, endoscopic ultrasound (EUS) guided fine needle aspiration (FNA), fine needle biopsy (FNB), single incision needle-knife biopsy (SINK), and endoscopic resection allow sampling of deeper tissue layers and may provide a histologic diagnosis. However there are no standard biopsy methods to diagnose SELs.

Endoscopic ultrasound (EUS) is a helpful imaging diagnostic tool for the evaluation of SELs as it is capable of assessing its size, layer of origin, and echo patterns. However, it is not reliable enough to differentiate between benign and malignant SELs. The combination of EUS and fine needle aspiration (FNA) allows for cytologic diagnostic accuracy of about 80%. If a malignant tumor is suspected, tissue acquisition is important because the management may vary based on the type of SEL.

EUS-FNA enables a small needle to be passed into the lesion of interest under ultrasound guidance, obtaining cellular material for cytology analysis. Although previous reports show accuracy rates for EUS-FNA for SEL that range between 72% and 90% there often is a problem of inadequate cellular aspirates for additional immunocytochemical examination. This problem causes a miss rate of approximately 40% for additional immunocytochemical analyses that may be needed to reach a definitive diagnosis for GISTs, although the macroscopic appearance alone is often sufficient. Using a larger FNA needle to perform the biopsy seems to be a logical solution to this limitation. However, doing so has thus far not been shown to improve the adequacy of biopsy aspirates.

For this reason, core biopsy needles (EUS-FNB) have been developed in the hopes of obtaining core tissue for histology. The advantages of a biopsy core specimen are well known because the evaluation of tissue architecture increase the diagnostic yield compared to cytology obtained from FNA. In addition, a tissue core biopsy is critical to diagnose and characterize SELs by providing higher rates of immunohistological evaluation. Recently, a novel design core biopsy needle (Sharkcore, Medtronic) has been introduced with preliminary data suggesting superior diagnostic performance compared to previous versions of core biopsy needles. How this new needle performs in the diagnostic evaluation of SELs remains unclear.

Single-incision needle-knife (SINK) biopsy is an alternative diagnostic method to acquire tissue samples. By using a conventional needle knife sphincterotome, the overlying mucosa of SELs is opened with a single linear incision of 10 mm and tissue samples are obtained by passing conventional biopsy forceps through the opening and deep into the tumor. According to a preliminary study by de la Serna Higuera et al, this technique provides sufficient tissue samples with high diagnostic yield of 92.8%. However, this biopsy method is only possible when a visible bulge from the SEL is apparent within the lumen of the gastrointestinal tract.

The investigators believe this technique offers the potential to obtain more substantial tissue samples that will be much more likely to provide a definitive diagnosis for subepithelial tumors. In addition, it is hoped that biopsy samples obtained via SINK will be better able to provide adequate tissue for determination of the mitotic rate of GIST tumors, which is a major predictor of malignant risk for these lesions. However, how the SINK method compares to the latest EUS-FNB core needles is unclear, both in terms of the diagnostic efficacy in obtaining the diagnosis and in terms of the safety profile of doing so.

The purpose of this study is to prospectively compare the efficacy and safety of EUS-FNB using the new SharkCore needle with SINK biopsy in patients with upper GI SELs. The results of the study will determine which method should be performed to confirm the diagnosis for SELs.

Inclusion Criteria:

• Undergoing diagnostic endoscopy for gastrointestinal SELs
• Tumor size ≥ 15 mm with endoscopically visible bulge
• SELs in esophagus, stomach, or duodenum
• Ability to sign the informed consent for diagnostic procedure

Exclusion Criteria:

• Lesions not requiring pathological evaluation (e.g. lipoma, cyst, varices)
• Underlying medical condition that contraindicates diagnostic endoscopy.
• Bleeding diathesis
• Inability to sign the informed consent.