BACKGROUND AND AIMS: The International Consensus Guidelines updated in 2017 recommended surgery to all main duct intraductal papillary mucinous neoplasms (MD-IPMNs) with the main pancreatic duct (MPD) of 10 mm or more and those with mural nodules regardless of size. The aim of the present study was to identify predictors of malignancy in MD-IPMN among preoperative factors including MPD and mural nodule size. METHODS: Twenty-six benign MD-IPMNs (7 resected and 19 nonresected) and 32 malignant MD-IPMNs (31 resected and 1 nonresected) were included in the study. MRCP, CT, EUS, and cytology were performed using pancreatic juice collected by endoscopic retrograde pancrestography (ERP). Resected IPMNs were classified as benign or malignant by histologic examination and nonresected MD-IPMNs by imaging, cytology, and observation. Cutoff values of candidate parameters were determined by receiver operating characteristic curves. Univariate and multivariate analyses by regression model were performed. RESULTS: MPD and mural nodule size and cytology results differed significantly between benign and malignant groups. Cutoff values of MPD and mural nodule sizes were 15 mm and 10 mm with areas under the curve of .66 and .86, respectively. Mural nodules of 10 mm or more (odds ratio, 8.32; 95% confidence interval, 1.13-61.2; P = .038) and positive cytology (odds ratio, 42.5; 95% confidence interval, 4.10-439; P = .002) were shown to be independent predictors of malignancy by multivariate analysis. When MD-IPMNs with either predictor were diagnosed to be malignant, sensitivities, specificities, and overall accuracy for malignancy were 94%, 85%, and 90%, respectively. CONCLUSIONS: Mural nodules of 10 mm or more and positive cytology were independent predictors of malignancy in MD-IPMN.
BACKGROUND AND AIMS: The International Consensus Guidelines updated in 2017 recommended surgery to all main duct intraductal papillary mucinous neoplasms (MD-IPMNs) with the main pancreatic duct (MPD) of 10 mm or more and those with mural nodules regardless of size. The aim of the present study was to identify predictors of malignancy in MD-IPMN among preoperative factors including MPD and mural nodule size. METHODS: Twenty-six benign MD-IPMNs (7 resected and 19 nonresected) and 32 malignant MD-IPMNs (31 resected and 1 nonresected) were included in the study. MRCP, CT, EUS, and cytology were performed using pancreatic juice collected by endoscopic retrograde pancrestography (ERP). Resected IPMNs were classified as benign or malignant by histologic examination and nonresected MD-IPMNs by imaging, cytology, and observation. Cutoff values of candidate parameters were determined by receiver operating characteristic curves. Univariate and multivariate analyses by regression model were performed. RESULTS: MPD and mural nodule size and cytology results differed significantly between benign and malignant groups. Cutoff values of MPD and mural nodule sizes were 15 mm and 10 mm with areas under the curve of .66 and .86, respectively. Mural nodules of 10 mm or more (odds ratio, 8.32; 95% confidence interval, 1.13-61.2; P = .038) and positive cytology (odds ratio, 42.5; 95% confidence interval, 4.10-439; P = .002) were shown to be independent predictors of malignancy by multivariate analysis. When MD-IPMNs with either predictor were diagnosed to be malignant, sensitivities, specificities, and overall accuracy for malignancy were 94%, 85%, and 90%, respectively. CONCLUSIONS: Mural nodules of 10 mm or more and positive cytology were independent predictors of malignancy in MD-IPMN.