S252 Endoscopic Ultrasound-Guided Biopsy of Mural Nodule in Patients With Intraductal Papillary Mucinous Neoplasm: A Viable Strategy to Guide Decision-Making?

Introduction: Intraductal papillary mucinous neoplasm with mural nodules (IPMN-MNs) are associated with a 30% to 50% risk of harboring high-grade dysplasia/adenocarcinoma (HGD/Ca). There are limited data on the accuracy of endoscopic ultrasound guided aspiration/biopsy (EUS-FNA) to identify HGD/Ca and its impact on management. This study evaluated the congruency of EUS-FNA results with surgical pathology. Methods: In a multi-center retrospective review, we identified patients with IPMN-MNs who underwent EUS-FNA between 2001 and 2024. Patient demographics, disease characteristics, and treatment modalities were summarized. Univariate analysis was performed to assess factors associated with false-negative EUS-FNA results. Kaplan-Meier analysis was performed to compare survival outcomes in patients with HGD/Ca between treatment groups. Results: We identified 239 patients; of whom majority (133, 56%) was men, with a median age of 74 (IQR 67-82). The median cyst size was 23 mm (IQR 16-33), and median mural nodule size was 8 mm (IQR 6-10) with a median follow-up of 4.4 years (IQR 1.7-8.5). Majority had branch duct IPMNs (161, 67%), followed by main duct IPMNs (42, 18%), and mixed-type IPMNs (36, 15%). Ninety-two (38%) underwent upfront resection, 30 (13%) underwent resection during surveillance, and 117 (49%) were managed with surveillance only. Among resected patients, the overall incidence of HGD/Ca on surgical pathology was 44%. Among 32 patients with an EUS-FNA result of HGD/Ca, 27 patients had a true-positive result (84% positive predictive value). Among 90 patients with EUS-FNA showing no HGD/Ca, 63 patients had a true-negative result (70% negative predictive value). Factors associated with false-negative EUS-FNA results were main or mixed duct involvement, pancreatic ductal dilatation ≥7 mm, and cyst size ≥33 mm on EUS (P < 0.05). EUS-FNA had a sensitivity of 50% (CI: 37%-63%); specificity of 93% (CI: 84%-97%); positive likelihood ratio of 6.8 (CI: 2.8-16.5) and negative likelihood ratio of 0.54 (CI: 0.4-0.7). Among patients with HGD/Ca on surgical pathology, overall survival was similar between patients managed with surveillance first before eventually undergoing resection and those who underwent upfront resection (P = 0.77). Conclusion: Among patients with IPMN-MN, 56% did not have HGD/Ca. Our study demonstrated that EUS-FNA of the mural nodule had a negative predictive value of 70% and may prove as a reasonable risk-stratifying modality especially in higher-risk patients with competing comorbidities.