The Natural History of Upper Gastrointestinal Subepithelial Tumors A Multicenter Endoscopic Ultrasound Survey Kanwar R. S. Gill, MD,* Lorenzo Camellini, MD,w Rita Conigliaro, MD,z Romano Sassatelli, MD,w Francesco Azzolini, MD,w Alessandro Messerotti, MD,z Timothy A. Woodward, MD,* Micheal B. Wallace, MPH, MD,* Laith H. Jamil, MD,* and Massimo Raimondo, MD* Goals: To evaluate the natural course of <3-cm upper gastro- intestinal subepithelial tumors by endoscopic ultrasound (EUS) and to determine the appropriate timing for EUS follow-up. Background: Subepithelial tumors (SETs) can range from benign lesions to tumors with malignant behavior or potential to become malignant such as gastrointestinal stromal tumors (GISTs). EUS is considered a valuable tool for their evaluation as it estimates the exact size and layer of origin, and also additional morphologic features that can suggest the diagnosis. For high surgical risk patients and when no worrisome EUS features are seen, EUS surveillance of subepithelial tumors is often used. Methods: Fifty-one patients (mean age, 61.2 ± 11.8 y; median, 63 y) with asymptomatic <3-cm SETs of second and fourth echolayer were followed for a mean period of 29.7 months (range, 3 to 84; median, 23 mo) in 3 tertiary care institutions. Evaluation included location, echolayer, tumor diameter, internal echo pattern, and outer margin of lesions by EUS. EUS was performed by using miniprobes, radial and linear echoendoscopes. Results: Follow-up revealed increase in size and/or change in echogenic features in 7/51 (13.7%) patients. Surgical follow-up was available for 3 of 7 of these patients. Two of the fourth layer SETs, which had both increase in size and change in echogenicity were found to be GISTs (+c-kit). Conclusions: The majority of <3-cm SETs does not change during a median of 23 months. The change in echogenicity and increase in size may indicate a GIST. Key Words: subepithelial tumors, endoscopic ultrasound (EUS), fine needle aspiration (FNA) (J Clin Gastroenterol 2009;43:723–726) S ubepithelial tumors (SETs) are relatively common findings in patients undergoing imaging with barium contrast radiography or esophagogastroduodenoscopy (EGD). SETs can range from benign lesions to tumors with malignant behavior or potential such as gastrointest- inal stromal tumors (GISTs) and carcinoids. 1,2 Endoscopic ultrasound (EUS) is considered a valuable tool for the evaluation of these tumors as it defines the layer of origin, closely estimates the size, and may show additional morphologic features for the diagnosis. The hypoechoic SETs arising from second echolayer (muscularis mucosa) or fourth echolayer (muscularis propria) on EUS examination can represent mesenchymal tumors such as GISTs. Among these, large (>3 cm) hypoechoic masses arising from the fourth echolayer (muscularis propria), lesions with in- homogeneous echogenicity, cystic spaces, and irregular margins that invades into other layers or structures on EUS examinations are more likely to represent malignant processes. 3,4 The exact natural history of the second echolayer hypoechoic subepithelial lesions is not well known. The EUS-guided fine needle aspiration (FNA) may provide tissue diagnosis on immunohistochemistry, like c-kit (CD117) staining for GISTs. 5,6 This is critical as all GISTs are considered to have malignant potential, and surgical resection is usually recommended. 7 However, for small (<3 cm) tumors arising from the fourth wall layer, tissue sampling by FNA may have low diagnostic yield. 8 In such uncharacterized SETs with no tissue diagnosis and absence of worrisome features on EUS (size <3 cm, homogenous, regular margins), the management is unclear. This is especially important in high surgical risk patients. For such patients, EUS surveillance of SET can be considered after discussion with patients and the under- standing that there is a risk of presence of GISTs, which may progress during the follow-up. 9 Nevertheless, the optimal management of these uncharacterized SETs re- mains unclear, and the ideal timing of endoscopic follow-up or EUS surveillance is not known. OBJECTIVES To evaluate the natural history of <3 cm hypoechoic upper gastrointestinal SETs arising from the second and fourth echolayer on EUS examination, and to determine the appropriate timing for EUS follow-up. MATERIALS AND METHODS We conducted a retrospective chart review of all EUS procedures performed from 1996 to 2007 for SETs at 3 tertiary care referral centers. Patients with at least 1 Copyright r 2009 by Lippincott Williams & Wilkins Received for publication March 21, 2008; accepted August 13, 2008. From the *Division of Gastroenterology and Hepatology, Mayo Clinic, Jacksonville, FL; wGastrointestinal Endoscopy Unit, S. Maria Nuova Hospital, Reggio Emilia; and zGastrointestinal Endoscopy Unit, New S. Agostino-Estense Hospital, Modena, Italy. Sources of Support: None. Disclaimers: All authors have no conflict of interest to declare. Reprints: Massimo Raimondo, MD, 4500 San Pablo Road, Mayo Clinic, Jacksonville, FL 32224 (e-mail: raimondo.massimo@ mayo.edu). ORIGINAL ARTICLE J Clin Gastroenterol  Volume 43, Number 8, September 2009 www.jcge.com | 723