The role of pancreatic juice cytology in the diagnosis of pancreatic intraductal papillary mucinous neoplasm.

Background And Aim: pancreatic juice cytology (PJC) is an important predictor of malignant intraductal papillary mucinous neoplasm (IPMN). This study aimed to determine the role of PJC for the prediction of malignant IPMN (intraductal papillary mucinous cancer [IPMC]).

Methods: medical records of IPMN patients who underwent surgery between 2012 and 2016 at the Nagasaki University Hospital were reviewed. Patients who underwent preoperative PJC were classified as high risk stigmata (HRS), worrisome features (WF) and no-criteria, based on imaging criteria. PJC class III or higher was considered as positive and only invasive IPMN was defined as IPMC. PJC was validated in each group with regard to sensitivity, specificity, accuracy with the corresponding 95% confidence intervals (95% CI) and area under receiver operating curve (AUROC) analysis. A p-value of < 0.05 was considered as statistically significant.

Results: preoperative pancreatic juice was obtained in 33/52 IPMN patients; only patients with adequate aspirate for cytology (n = 29) were included. In the HRS group (n = 9), 4/6 non-IPMC had a negative PJC and 3/3 IPMC had a positive PJC. In the WF group (n = 17), 9/11 non-IPMC had a negative PJC and 3/6 IPMC had a positive PJC. Adding PJC to imaging results improved the AUROCs of HRS and WF from 0.63 and 0.62 to 0.83 and 0.66, respectively. PJC was negative in all no-criteria cases (n = 3; one IPMC and two non-IPMC). In all 29 patients, PJC sensitivity was 60% (95% CI: 26%-88%), specificity was 79% (95% CI: 54%-94%), accuracy was 72% (95% CI: 63%-89%) and the AUROC was 0.69 (p = 0.03).

Conclusion: PJC is a statistically significant IPMC predictor that can improve the validity of imaging for IPMC prediction.