CYSTIC PANCREATIC LESIONS: IMAGING VERSUS ANATOMOPATHOLOGICAL FINDINGS-HOW TO IMPROVE DIAGNOSTIC ACCURACY?

Background: Pancreatic cystic lesions are a group of pancreatic neoplasms with different behavior and risk of malignancy. Imaging diagnosis and differentiation of these lesions remain a challenge.
Aims: The aim of this study was to evaluate the agreement between computed tomography and/or magnetic resonance imaging and post-operative pathologic diagnoses of Pancreatic cystic lesions in a University Hospital of São Paulo State.
Methods: A total of 39 patients with surgically diagnosed Pancreatic cystic lesions were enrolled, as a study cohort from 2009 to 2019. Preoperative radiological and final pathological diagnosis was correlated to measure computed tomography and/or magnetic resonance imaging diagnostic. Pancreatic adenocarcinoma, choledochal pancreatic cyst, mucinous cystadenoma, serous cystadenoma, intraductal papillary mucinous neoplasms, and pancreatic pseudocyst were classified as neoplastic cysts.
Results: It was noted that 27 patients (69.23%) had preoperative computed tomography and magnetic resonance imaging, 11 patients (28.20%) had preoperative computed tomography only, and 1 patient had preoperative magnetic resonance imaging only. The values for diagnoses made only with computed tomography (p=0.47) and from the combination of computed tomography+magnetic resonance imaging (p=0.50) did also point to moderate agreement with the anatomopathological findings. The values pointed to a fair agreement for the diagnosis of mucinous cystadenoma (p=0.3), moderate agreement for intraductal papillary mucinous neoplasms (p= 0.41), good agreement for serous cystadenoma (p=0.79), and excellent agreement for choledochal pancreatic cyst (p=1), pancreatic pseudocyst (p=0.84), and Frantz tumor (p=1) (p<0.05).
Conclusions: The findings of computed tomography and/or magnetic resonance imaging have an equivalent diagnostic agreement with an anatomopathological diagnosis for differentiating benign from malignant Pancreatic cystic lesions and in suggesting a specific diagnosis. There is no statistical difference between the use of computed tomography alone and computed tomography+magnetic resonance imaging in the improvement of diagnostic accuracy.