Malignant intraductal papillary mucinous neoplasm of the pancreas: In situ versus invasive carcinoma-surgical resectability

Purpose:To retrospectively evaluate computed tomographic ( CT) findings in patients with in situ and invasive malignant intraductal papillary mucinous neoplasms ( IPMNs) of the pancreas and to evaluate the accuracy for surgical resectability, with surgery and pathologic analysis as the reference standards. Materials and Methods: Institutional review board approval was obtained, and the informed consent requirement was waived. Forty- six patients with malignant IPMN proved at pathologic examination of the surgically resected specimen ( n = 44) or laparotomy ( n = 2) underwent surgery after multidetector CT was performed. CT findings were retrospectively evaluated to determine if a pancreatic malignant IPMN tumor was present; to make this determination, CT criteria were used to differentiate in situ from invasive tumors and signs of unresectability ( liver metastasis, vascular CT pattern of encasement, or regional lymph node metastasis). The extent of the vascular CT pattern of encasement was recorded for each patient ( no obliteration of the fat plane, obliteration of the fat plane of < 50%, or obliteration of the fat plane of >= 50%). Statistical analysis was performed with the x 2 and Student t tests. Results: CT revealed a mural nodule in the pancreatic duct wall in 14 patients with in situ carcinoma and one patient with invasive carcinoma ( P <.003). CT revealed an infiltrative pancreatic mass in 17 patients with invasive carcinoma and two patients with in situ carcinoma ( P <.02). Of the mural nodules, 93% were seen in patients with in situ carcinoma, whereas 90% of infiltrative pancreatic masses were observed in patients with invasive carcinomas. The positive predictive value of CT for determining resectability was 100%, and the overall accuracy of CT for determining resectability and unresectability was 74%. The positive predictive value of CT for determining unresectability was 17%, mainly owing to overestimation of arterial invasion. Conclusion: CT is helpful in the differentiation of in situ and invasive IPMN. Classic vascular invasion criteria lead to the overestimation of surgical tumor unresectability in patients with malignant IPMN.