Imaging Features of Computed Tomography of Non-benign Intraductal Papillary Mucinous Neoplasm of Pancreas

  Objective  To evaluate the imaging features of computed tomography (CT) of non-benign intraductal papillary mucinous neoplasm (IPMN) of pancreas, and to explore the correlation between radiological and pathological finding.

  Methods  CT images of 12 patients who were pathologically diagnosed as non-benign IPMN of pancreas between May, 2008 and February, 2011 in Peking Union Medical College Hospital were retrospectively evaluated. Their pre-surgery CT images were reviewed by 2 experienced radiologists, and the radiopathological results were compared to evaluate the diagnostic value of CT techniques.

  Results  Of these 12 patients, 4 were pathologically proved as borderline IPMN, 2 as non-invasive intraductal papillary mucinous carci-noma (IPMC), and 6 as invasive IPMC. The clinical manifestations including weight loss, jaundice and diabetes suggested malignancy, while abdominal discomfort or pain has no specificity. IPMN were divided into 3 types according to CT classification, namely main pancreatic duct type, branch pancreatic duct type and mixed type. The branch duct type were rarely seen in malignant IPMN cases. The common radiological features of IPMN included pancreatic duct dilation or large cysts formation (with septa or focal nodules inside), pancreatic parenchymal calcification and atropy, which were seen in both borderline IPMN and IPMC. However, for malignant cases, the pancreatic duct dilation were much severer; CT values in the dilated pancreatic ducts were higher, and showed a heterogeneity within the duct. Common bile duct dilation were seen in invasive IPMC only.

  Conclusions  Multi-slice spiral CT is valuable in the pre-surgery diagnosis of non-benign IPMN, and shows a good agreements with pathological results. The CT features of IPMN lesions can predict the degree of malignancy and the involvement of surrounding tissues at a favorable accuracy, and therefore provides useful information for clinical diagnosis and decision for surgery.