AbstractBackgroundWe compared the results of preoperative pancreatic juice cytology (PJC) and final pathological diagnosis after resection in patients who underwent resection of intraductal papillary mucinous neoplasm (IPMN) of the pancreas to determine whether preoperative PJC can help determine therapeutic strategies.MethodsOf 1130 patients who underwent surgical resection IPMN at 11 Japanese tertiary institutions, the study included 852 patients who underwent preoperative PJC guided by endoscopic retrograde cholangiopancreatography (ERCP).ResultsThe accuracy of preoperative PJC for differentiation between cancerous and noncancerous lesions were 55% for IPMN overall; 59% for the branch duct type; 49% for the main pancreatic duct type; 53% for the mixed type, respectively. On classifying IPMN according to the diameters of the mural nodule (MN) and main pancreatic duct (MPD), the corresponding values for diagnostic performance were 40% for type 1 (MN ≥5 mm and MPD ≥ 10 mm); 46% for type 2 (MN ≥5 mm and MPD < 10 mm); 61% for type 3 (MN < 5 mm and MPD ≥ 10 mm); 72% for type 4 (MN < 5 mm and MPD < 10 mm), respectively.ConclusionsPJC in IPMN is not a recommended examination because of its low overall sensitivity and no significant difference in diagnostic performance by type, location, or subclassification. Although the sensitivity is low, the positive predictive value is high, so we suggest that pancreatic juice cytology be performed only in cases where the patient is not sure about surgery.