Endoscopic ultrasound and endoscopic ultrasound-guided fine needle aspiration biopsy in the evaluation of upper gastrointestinal tract subepithelial lesions – a single center’s experience in the Philippines
Background: Subepithelial masses (SEMs) are common incidental findings during routine endoscopy and endoscopic ultrasound (EUS) has been established as the modality of choice for the evaluation of these lesions. Diagnostic accuracy for differentiating benign versus malignant SEMs range from 48-68% in some studies. In the Philippines, EUS has only been recently used for the past 5 years and is limited to only a few centers nationwide. This study created a descriptive profile of our center’s pioneering EUS database and determined the center’s diagnostic accuracy.
Methods: All patients referred to the Institute of Digestive and Liver Diseases (IDLD) of St. Luke’s Medical Center-Global City from January 2010-December 2014 who underwent upper gastrointestinal EUS and/or EUS-FNAB were retrospectively collected from an electronic database. The procedures were independently done by 6 trained endosonographic endoscopists. Intramural SEMs were classified based on the endosonographic features as assessed by each endoscopist into neoplastic or benign masses. Diagnostic accuracy was confirmed via EUS-FNAB. Incomplete EUS reports and studies on the pancreas & gallbladder were excluded.
Results: A total of 212 reports were included consisting of upper gastrointestinal SEMs in the oesophagus (42), stomach (94) and duodenum (37). A mass visualised on prior endoscopic study (80%) and subsequent follow-up evaluations (11%) were the most common indication for EUS referral. Overall diagnostic accuracy for neoplastic lesions was 79%, where EUS sonographic findings correctly diagnosed 47 out of 59 histologically confirmed lesions. The reclassified neoplastic SEMs were non-specific chronic inflammation and cysts. The most common neoplastic SEMs were gastrointestinal stromal tumours (40%) followed by adenocarcinomas (12%) and squamous cell carcinomas (7%). For benign lesions, only 8% were reclassified as neoplastic after histologic confirmation, which included a round cell malignancy
Conclusion: Our center was able to achieve remarkable diagnostic accuracy for EUS differentiation between neoplastic and benign SEMs compared to prior studies. We recommend a more robust prospective trial for all patients undergoing EUS and a standardised list of parameters to be included in each EUS form that will allow uniform assessment and minimise operator bias. This study will provide a standard specific for local ethnicity and an algorithm which can be implemented nationwide.