ENDOSCOPIC ULTRASOUND-GUIDED FINE-NEEDLE ASPIRATION FOR SUBEPITHELIAL LESIONS: A SINGLE CENTRE EXPERIENCE

Aims Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) is commonly used for obtaining tissue from gastrointestinal subepithelial lesions (SEL). However, the diagnostic accuracy of this technique ranges widely (46%-93%) depending on multiple factors. The aim of our study was to evaluate diagnostic yield of EUS-FNA of SEL and factors related to sampling adequacy.

Methods Single centre retrospective study including consecutive patients with SEL submitted to EUS-FNA from January/2012 to October/2019. The final diagnosis was based on clinical and imaging follow-up and/or pathology. Data were collected from electronic medical reports.

Results A total of 44 patients were included (male: 52.3%; median age: 76 (59.7-74.7 years). The majority of patients (59.1%) were asymptomatic. Most lesions were gastrointestinal stromal tumours (45.5%) and carcinomas (13.6%), were mostly located at stomach (47.7%) and duodenum (22.7%) and were mostly originated from muscularis propria (54.5%) and muscularis mucosa (11.4%). The median size of lesions was 30.0 (19.9-50.0) mm, a 22G needle was used in 90.9% of cases and the median number of passes was 3 (2.3-4.0). Sensitivity, specificity, positive and negative predictive values and diagnostic accuracy for malignancy were 71.4%, 77.8%, 100%, 63.6% and 72.7%, respectively. No adverse events were documented. Esophageal or gastric localizations (vs. duodenum and rectum) was predictive for sampling inadequacy (OR: 7.56; 95% CI: 1.08-52.9) while no differences were found between different needle size (22G vs. 25G: 70% vs. 100%; p = 0.30); number of passes (≤3 vs. >3: 78.1% vs. 50.0%; p = 0.07); lesion size (≤ 20 mm vs. >20 mm: 50.0% vs. 78.1%; p = 0.08); origin layer (muscularis mucosa vs. submucosa and muscularis propria: 90% vs. 64.7%; p = 0.12).

Conclusions Our diagnostic yield is similar to reported in the literature, making EUS-FNA an adequate method for obtaining tissue from SEL. However, it is slightly lower than the standard defined by European Society of Gastrointestinal endoscopy performance guidelines. Lesion location was the only predictor for sampling adequacy.