Nomogram for the Prediction of High-Grade Dysplasia and Invasive Carcinoma in Patients With Intraductal Papillary Mucinous Neoplasms of the Pancreas Based on Variables of Noninvasive Examination

Intraductal papillary mucinous neoplasms (IPMNs) represent the most common identifiable precursor lesions of pancreatic cancer. They are a heterogenous group of neoplasms, and clinical decision-making of malignant disease, including high-grade dysplasia and invasive carcinoma, is challenging. Moreover, discrepancy on the indication for resection exists in the current guidelines. Most of the current nomogram models for predicting malignant disease depended on endoscopic ultrasonography for the precise size of mural nodules. Thus, this study aims to propose a model to predict a malignant disease using variables from a noninvasive examination, including cross-sectional imaging and blood testing.We evaluated patients who underwent resection of pathologically confirmed IPMNs and had preoperative clinical data available for review from November 2010 to December 2018. Based on a binary logistic regression model, we devised a nomogram model to predict malignant IPMNs. The area under the receiver operating characteristics curve (AUC) was used to evaluate the discrimination power of the model. Of the 333 patients who underwent resection of IPMNs, 198 (59.5%) had benign IPMNs and 135 (40.5%) had malignant IPMNs. Binary logistic regression analysis showed that cyst size, cyst location, cyst wall enhancement, multicystic lesion, diameter of main pancreatic duct, neutrophil-to-lymphocyte ratio, serum carbohydrate antigen 19-9, and carcinoembryonic antigen were significantly associated with malignancy and thus were selected to construct the nomogram. The nomogram showed excellent discrimination power with an AUC of 0.859 (95% CI: 0.818-0.900, P<0.001). In this study, a combination of cross-sectional imaging features and blood markers was optimal in predicting malignancy and was incorporated into a nomogram model that was developed to predict malignancy. The nomogram, which uses variables of noninvasive examinations during the surveillance, was able to distinguish benign from malignant IPMNs. Nevertheless, external validation is warranted.