Endoscopic Unroofing for Small Gastric SETs

INTRODUCTION

Small gastric subepithelial tumors (SETs) are often incidental findings at upper endoscopy. Differential diagnoses include a variety of different pathologies such as gastrointestinal stromal tumors (GISTs), leiomyomas, leiomyosarcomas, lipomas, carcinoid tumors, granular cell tumors, duplication cysts or pancreatic rests. Since some of these entities have a potential of malignancy, accurate classification of gastric SETs is essential to define the best diagnostic and therapeutic management. Endosonography (EUS) is the current gold standard for further characterization of small gastric SETs. It provides differentiation between extra- and intramural lesions, information on the involvement of certain wall layers, measurement of a lesion as well as structural analysis according to the echogenicity of a lesion. Nevertheless, a level of uncertainty remains, even if a lesion shows a typical echo-pattern and layer involvement.

According to current guidelines, lesions of more than two centimeters in size that are suspected to be GISTs should be removed by surgical resection. Lesions less than one centimeter can be followed by annual EUS. For lesions bigger than one centimeter but smaller than two centimeters tissue sampling is recommended in order to correctly diagnose a lesion according to immunohistochemical markers (such as CD117 or CD34 in GISTs) and to quantify the malignant potential according to other histological factors (such as the mitotic index in GISTs) [1].

Tissue sampling can be done by using standard or jumbo forceps biopsy and EUS-guided fine needle aspiration or biopsy, but the diagnostic yield of these methods is rather low [2,3]. Better results can be achieved by endoscopic resection techniques such as endoscopic submucosal dissection (ESD), submucosal tunneling endoscopic resection or full thickness endoscopic resection [4,5]. The main advantage of these approaches is that they are not only diagnostic but also therapeutic strategies since the respective lesion can be resected as a whole comparable with conventional surgery. However, there is a high risk of procedure related adverse events such as bleeding or perforation (the latter is often unavoidable considering that most SETs are arising from deep wall layers such as the muscularis propria layer).

Endoscopic unroofing is very similar to the mentioned endoscopic resection techniques with the main difference that only partial resection is intended in order to get a representative tissue specimen for histological evaluation. This method has been described in a case series by Lee et al. in 2010 where it showed a very favorable diagnostic yield in gastric SETs [6]. To date, no study has reported on the follow-up of lesions investigated by endoscopic unroofing.

The aim of the present study is to evaluate endoscopic unroofing for small gastric SETs regarding safety, diagnostic yield and potential therapeutic effects over time.

PRIMARY HYPOTHESIS

Endoscopic unroofing has a high diagnostic accuracy in small gastric subepithelial tumors.

METHODS

This retrospective study will be conducted at the Division of Gastroenterology and Hepatology, Department of Internal Medicine III. Of the Medical University of Vienna. Patients who underwent endoscopic unroofing of gastric subepithelial tumors between 2000 and 2015 will be identified by searching the endoscopy reporting database "DOKMON". Clinical data will be collected from patients identified by reviewing endoscopy examination reports as well as other clinical documents including discharge letters in the report archiving database "AKIM". The following parameters will be assessed: Demographics (age, sex), indication for endoscopic unroofing, intraprocedural adverse events, hospital stay, histological result, follow-up procedures (including endoscopic or imaging controls, surgical resection). At the end of the study the mentioned parameters will be presented using descriptive statistics. Data analysis will be performed using SPSS 23.0. All patient related data will be de-identified using pseudonymization prior to any further processing.

SAMPLE SIZE

The investigators expect to encounter approximately 20 patients into this study (between 2000 and 2015).