S3239 Evaluation of Sensitivity and Diagnostic Accuracy of Endoscopic Ultrasound for Diagnosis of Solid Pancreatic Lesions

Abstract

INTRODUCTION: Endoscopic ultrasonography with fine needle aspiration/biopsy (EUS-FNA/FNB) facilitates a minimally invasive approach to obtaining pancreatic tissue for the workup of possible malignancies. The American Society of Gastrointestinal Endoscopy (ASGE) has set the benchmark for EUS-FNA/FNB to have a diagnostic rate of >75% and a sensitivity of >85% for solid pancreatic masses. The aim of this study is to assess the real-world rates of diagnostic accuracy, sensitivity and specificity of EUS-FNA for solid pancreatic masses, and factors associated with higher accuracy and sensitivity. METHODS: We conducted a retrospective chart review of all adult patients who underwent EUS-FNA/FNB at an academic level-1 trauma center in southeastern New England between April 2015 and March 2020. Patients were selected based on CPT billing codes for EUS with FNA/FNB. The primary outcomes were diagnostic accuracy and sensitivity, and these were stratified based on several intraprocedural factors including tumor size and number of passes. Samples with pathology reports of atypia without confirmed malignancy were considered non-malignant. RESULTS: Of 204 patients who underwent EUS with FNA/FNB, 38 were done for solid pancreatic masses. 166 patients were excluded due to cystic lesions (49) and 117 were done for intramural lesions or EUS-liver biopsy. For solid masses, the mean maximum tumor diameter was 26.9mm, with 46% occurring in the head or uncinate process (Table 2). The mean number of passes was 2.39 and 87% were done with a 22-gauge needle. On-site cytopathology was present for 37% of cases and 21% were done with general anesthesia. 53% had a duodenal approach. The overall diagnostic rate was 95%. Sensitivity for pancreatic malignancy was 85% and specificity was 100%. Sensitivity and diagnostic rates are higher with increasing tumor size, 2-4 needle passes vs 2-3 passes, and for masses in the pancreatic tail vs neck (Table 2). CONCLUSION: he diagnostic accuracy and sensitivity of EUS-FNA/FNB for solid pancreatic lesions were 95% and 85%, respectively, at our institution, thus meeting the ASGE grade 1C guidelines, despite fewer passes and less on-site cytopathology presence compared with larger studies[2].Table 1.: Demographics of patients with solid pancreatic tumorsTable 2.: Subgroup analysis by needle size, number of passes, tumor stage and approach (trans-gastric vs. trans-duodenum)