ABCD: Arquivos Brasileiros de Cirurgia Digestiva ()
MAGNETIC RESONANCE CHOLANGIOPANCREATOGRAPHY (MRCP) VERSUS ENDOSONOGRAPHY-GUIDED FINE NEEDLE ASPIRATION (EUS-FNA) FOR DIAGNOSIS AND FOLLOW-UP OF PANCREATIC INTRADUCTAL PAPILLARY MUCINOUS NEOPLASMS
Abstract
ABSTRACT Background: Intraductal papillary mucinous tumor (IPMN) are being diagnosed with increasing frequency. Computerized tomography scanning is commonly used as the primary imaging modality before surgery nonetheless magnetic resonance cholangiopancreatography (MRCP) provides better characterization. Endosonography-guided fine needle aspiration (EUS-FNA) has emerged as a way to reach pathological diagnose. Aim: To compare results of both methods with surgical pathology findings for classification of IPMN. Methods: Thirty-six patients submitted to surgical resection with preoperative suspect of IPMN were submitted preoperatively to MRCP and EUS-FNA. Images obtained were analyzed according to a classification determined for each method. ROC curve was used for statistical analysis, that compared the images tests with the purpose of finding the best method for diagnosis and classification of IPMN. Results: Sixteen patients underwent pancreatoduodenectomy, 16 to subtotal pancreatectomy and only four laparotomy. Pathological diagnosis was IPMN (n=33) and pancreatic intraepithelial neoplasia type 2 (n=3). Twenty-nine revealed non-invasive neoplasia and invasive form in four patients. MRCP and EUS-FNA have correctly diagnosed and classified (type of IPMN), in 62.5% and 83.3% (p=0.811), the affected segment location in 69% and 92% (p=0.638) and identification of nodules and/or vegetation presence in 45% and 90% (p=0.5). Regarding to histopathological diagnosis by EUS-FNA the sensitivity was 83.3%; specificity was 100%; positive predictive value was 100%; negative predictive value was 33.3% and accuracy was 91.7%. Conclusions: There was no significant difference in the diagnosis of IPMN. However, EUS-FNA showed better absolute results than MRCP to identify nodule and/or vegetation.
Keywords