[Characterization and diagnostic accuracy of serous cystadenomas and mucinous neoplasms of the pancreas with multi-slice helical computed tomography].

OBJECTIVE: To evaluate the feasibility of using multi-slice helical computed tomography (MS-CT) to accurately distinguish serous cystadenomas from mucinous cystadenomas or cystadenocarcinomas of the pancreas and to determine their radiographic appearances that can be applied for differentiative diagnosis. METHODS: We performed a single-blind retrospective analysis of CT images of 30 patients with pathologically proven primary cystic pancreatic neoplasms (12 cases of serous cystadenomas, 14 cases of mucinous cystadenomas, and 4 cases of mucinous cystadenocarcinomas) to reach a diagnosis of either serous cystadenoma or mucinous cystic tumor. CT features such as tumor location, septations, presence of calcification, features of cystic wall, papillary excrescences, and size of the largest cyst were recorded. Statistical analysis was performed to evaluate the efficacy of certain CT findings in the differentiation of serous cystadenomas and mucinous neoplasms. RESULTS: Totally 9 (75.0%) serous cystadenomas and 16 (88.9%) mucinous tumors were correctly diagnosed. Three serous cystadenomas were misdiagnosed as mucinous cystadenomas, while 2 mucinous neoplasms were misdiagnosed as serous cystadenomas. And 9 (75.0%) serous cystadenomas were located at the pancreatic head and neck areas, while 12 (66.7%) mucinous neoplasms were located at the pancreatic body and tail areas (P < 0.05). The presence of calcification, especially central calcification, had statistical significance in differentiating serous cystadenoma from mucinous neoplasms (P < 0.05). The size of the largest cyst over 2 cm was positive associated with mucinous neoplasms (P < 0.05). CONCLUSION: CT characteristics between serous cystadenomas and mucinous neoplasms of the pancreas have distinct difference, which validates the values of CT in differentiating these tumors. However, atypical CT appearances may compromise its diagnostic accuracy.