Yield and performance characteristics of endoscopic ultrasound-guided fine needle aspiration for diagnosing upper GI tract stromal tumors.

BACKGROUND AND AIMS: EUS-FNA is a means of sampling suspected GI stromal tumors (GIST). However, there are limited published data on factors influencing the sampling yield, and on the performance characteristics of this technique in comparison with resection pathology. We analyzed the yield of EUS-FNA for submucosal lesions of the upper GI tract, and determined the performance characteristics of EUS-FNA for diagnosing GISTs. METHODS: We retrospectively reviewed procedural and pathology data from consecutive patients undergoing EUS-FNA of submucosal lesions from two medical centers over a 4-year period. We analyzed the yield of EUS-FNA, and calculated performance characteristics of EUS-FNA for GIST based on resection pathology. RESULTS: A total of 65 patients underwent EUS-FNA of 66 submucosal lesions during the study period. EUS-FNA was either diagnostic (68%) or suspicious (12%) in a total of 80%. EUS-FNA yielded the following diagnoses: GIST based on cytology and immunohistochemistry (56%), suspected GIST (12%), leiomyoma (9%), other neoplasm (3%), and non-diagnostic (20%). Larger lesion size, gastric location, and presence of on-site cytopathology were associated with higher yield in univariate analysis. Larger needle size and number of FNA passes were not associated with improved yield. Based on resection pathology from 28 specimens, the EUS-FNA performance characteristics for diagnosing GISTs included a sensitivity of 82%, a specificity of 100%, and an overall accuracy of 86%. CONCLUSIONS: EUS-FNA provides a high yield for sampling submucosal lesions and is highly accurate for diagnosing GISTs. EUS-FNA has an important role in the evaluation of suspected GISTs.