Secondary Logo

Journal Logo

Institutional members access full text with Ovid®

New Model for Predicting Malignancy in Patients With Intraductal Papillary Mucinous Neoplasm

Shimizu, Yasuhiro MD*; Hijioka, Susumu MD; Hirono, Seiko MD; Kin, Toshifumi MD§; Ohtsuka, Takao MD; Kanno, Atsushi MD||; Koshita, Shinsuke MD**; Hanada, Keiji MD††; Kitano, Masayuki MD‡‡; Inoue, Hiroyuki MD§§; Itoi, Takao MD¶¶; Ueki, Toshiharu MD||||; Matsuo, Keitaro MD***; Yanagisawa, Akio MD†††; Yamaue, Hiroki MD; Sugiyama, Masanori MD‡‡‡; Okazaki, Kazuichi MD§§§

doi: 10.1097/SLA.0000000000003108
Original Article: PDF Only

Objective: To create a simple, objective model to predict the presence of malignancy in patients with intraductal papillary mucinous neoplasm (IPMN), which can be easily applied in daily practice and, importantly, adopted for any lesion types.

Background: No predictive model for malignant IPMN has been widely applied in clinical practice.

Methods: The clinical details of 466 patients with IPMN who underwent pancreatic resection at 3 hospitals were retrospectively analyzed for model development. Then, the model was validated in 664 surgically resected patients at 8 hospitals in Japan.

In the preoperative examination, endoscopic ultrasonography (EUS) was considered to be essential to observe mural nodules in both the model development and external validation sets. Malignant IPMNs were defined as those with high-grade dysplasia and associated invasive carcinoma.

Results: Of the 466 patients, 258 (55%) had malignant IPMNs (158 high-grade dysplasia, 100 invasive carcinoma), and 208 (45%) had benign IPMNs. Logistic regression analysis resulted in 3 variables (mural nodule size, main pancreatic duct diameter, and cyst size) being selected to construct the model. The area under the receiver operating characteristic curve (AUC) for the model was 0.763. In external validation sets, the pathological diagnosis was malignant and benign IPMN in 351 (53%) and 313 (47%) cases, respectively. For the external validation, the malignancy prediction ability of the model corresponded to an AUC of 0.725.

Conclusion: This predictive model provides important information for physicians and patients in assessing an individual's risk for malignancy and may help to identify patients who need surgery.

*Department of Gastroenterological Surgery, Aichi Cancer Center Hospital, Nagoya, Japan

Department of Hepatobiliary and Pancreatic Oncology, National Cancer Center, Tokyo, Japan

The Second Department of Surgery, Wakayama Medical University, School of Medicine, Wakayama, Japan

§Center for Gastroenterology, Teine-Keijinkai Hospital, Sapporo, Japan

Department of Surgery and Oncology, Graduate School of Medical Sciences, Kyushu University, Fukuoka, Japan

||Division of Gastroenterology, Tohoku University Graduate School of Medicine, Sendai, Japan

**Department of Gastroenterology, Sendai City Medical Center, Sendai, Japan

††Department of Gastroenterology, JA Onomichi General Hospital, Hiroshima, Japan

‡‡Department of Gastroenterology and Hepatology, Kindai University School of Medicine, Osakasayama, Japan

§§Department of Gastroenterology and Hepatology, Mie University Graduate School of Medicine, Tsu, Japan

¶¶Department of Gastroenterology and Hepatology, Tokyo Medical University, Tokyo, Japan

||||Department of Gastroenterology, Fukuoka University Chikushi Hospital, Fukuoka, Japan

***Division of Molecular Medicine, Aichi Cancer Center Research Institute, Nagoya, Japan

†††Department of Pathology, Kyoto Prefectural University of Medicine, Kyoto, Japan

‡‡‡Department of Surgery, Kyorin University School of Medicine, Tokyo, Japan

§§§Department of Gastroenterology and Hepatology, Kansai Medical University, Osaka, Japan.

Reprints: Yasuhiro Shimizu, MD, Department of Gastroenterological Surgery, Aichi Cancer Center Hospital, Kanokoden 1-1, Chikusa-ku, Nagoya 464-8681, Japan. E-mail:

This work was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Number 16K09403.

This study was supported by the Japan Pancreas Society. The design and conduct of the study, interpretation of the data, and decision to submit the manuscript for publication were the responsibility of the authors listed.

The authors have no conflicts of interest.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.