S0106 Imaging Resource Utilization in the Surveillance of Presumed Branched Duct Intraductal Papillary Mucinous Neoplasm (BD-IPMN)

Abstract

INTRODUCTION: Presumed mucinous cystic lesions are at increased (albeit low) risk of malignant transformation and require regular imaging for ongoing surveillance. A cost-effective evaluation of imaging resource utilization (IRU) is understudied due to a paucity of actual clinical utilization data and lack of consensus among current guidelines. We assessed the value proposition of IRU in surveillance of BD-IPMN, by defining populations utilizing high volumes of imaging, and identifying the predictors of high IRU. METHODS: This is a single-center cross-sectional study, of patients with presumed BD-IPMN under active surveillance as of January 2018. Patients were identified from a prospectively maintained pancreatic cyst neoplasm (PCN) database. Presumed diagnosis of mucinous cysts with identification of worrisome features (WF) and high-risk stigmata (HRS), was made based on the accepted criteria from the literature. IRU was calculated based on both noninvasive (CT/MRI) and minimally invasive (EUS) imaging used per month during surveillance. For analysis, the “High Resource Utilization” (HRU) group is defined as the top 10% of the population consuming higher imaging resources and compared them with rest of the surveillance population: “Low Resource Utilization” (LRU) group. The clinical variables were compared using chi-square analysis and student’s t test between two groups. RESULTS: 409 patients with presumed BD-IPMNs were followed for 41.60 ± 32.24 months. By definition, the rate of surveillance imaging utilization per month in HRU group was higher, 0.74 ± 0.42 vs 0.16 ± 0.09 (P < 0.001). Similarly, the rate of non-invasive imaging and EUS utilization was also higher in HRU group. The HRU group have a higher initial cyst size (18.33 ± 14.31 mm vs. 12.90 ± 9.08 mm) (P < 0.001), higher proportion of cysts with WF (17.5% vs 13.82%) (P < 0.001), or with HRS (7.5% vs 1.36%) (P < 0.001), compared with LRU group. Cyst growth rate per year (30.00% vs 4.88% (P < 0.001)) and per two years (35.00% vs 13.82% (P < 0.001)) displayed higher proportions of fast-growing cysts (>5 mm threshold) in the HRU group compared with LRU group. CONCLUSION: Higher IRU was found in a subset of presumed BD-IPMN population undergoing active surveillance using actual prospective clinical data. Predictors of this included larger initial cyst size, rapid cyst growth rate over one and two years, and the presence of WF and HRS. Further refinement of imaging guidelines is needed for accurate long term PCN surveillance balanced by cost-effective IRU.Table 1