Proposed Nomogram Predicting the Individual Risk of Malignancy in the Patients with Branch Duct Type Intraductal Papillary Mucinous Neoplasms of the Pancreas

Jin Young Jang, Taesung Park, Selyeong Lee, Yongkang Kim, Seung Yeoun Lee, Sun Whe Kim, Song Cheol Kim, Ki Byung Song, Masakazu Yamamoto, Takashi Hatori, Seiko Hirono, Sohei Satoi, Tsutomu Fujii, Satoshi Hirano, Yasushi Hashimoto, Yashuhiro Shimizu, Dong Wook Choi, Seong Ho Choi, Jin Seok Heo, Fuyuhiko MotoiIppei Matsumoto, Woo Jung Lee, Chang Moo Kang, Ho Seong Han, Yoo Seok Yoon, Masayuki Sho, Hiroaki Nagano, Goro Honda, Sang Geol Kim, Hee Chul Yu, Jun Chul Chung, Yuichi Nagakawa, Hyung Il Seo, Hiroki Yamaue

Research output: Contribution to journalArticlepeer-review

40 Citations (Scopus)

Abstract

Objectives: This study evaluated individual risks of malignancy and proposed a nomogram for predicting malignancy of branch duct type intraductal papillary mucinous neoplasms (BD-IPMNs) using the large database for IPMN. Background: Although consensus guidelines list several malignancy predicting factors in patients with BD-IPMN, those variables have different predictability and individual quantitative prediction of malignancy risk is limited. Methods: Clinicopathological factors predictive of malignancy were retrospectively analyzed in 2525 patients with biopsy proven BD-IPMN at 22 tertiary hospitals in Korea and Japan. The patients with main duct dilatation >10 mm and inaccurate information were excluded. Results: The study cohort consisted of 2258 patients. Malignant IPMNs were defined as those with high grade dysplasia and associated invasive carcinoma. Of 2258 patients, 986 (43.7%) had low, 443 (19.6%) had intermediate, 398 (17.6%) had high grade dysplasia, and 431 (19.1%) had invasive carcinoma. To construct and validate the nomogram, patients were randomly allocated into training and validation sets, with fixed ratios of benign and malignant lesions. Multiple logistic regression analysis resulted in five variables (cyst size, duct dilatation, mural nodule, serum CA19-9, and CEA) being selected to construct the nomogram. In the validation set, this nomogram showed excellent discrimination power through a 1000 times bootstrapped calibration test. Conclusion: A nomogram predicting malignancy in patients with BD-IPMN was constructed using a logistic regression model. This nomogram may be useful in identifying patients at risk of malignancy and for selecting optimal treatment methods. The nomogram is freely available at http://statgen.snu.ac.kr/software/nomogramIPMN.

Original languageEnglish
Pages (from-to)1062-1068
Number of pages7
JournalAnnals of surgery
Volume266
Issue number6
DOIs
Publication statusPublished - 2017 Dec 1

All Science Journal Classification (ASJC) codes

  • Surgery

Fingerprint Dive into the research topics of 'Proposed Nomogram Predicting the Individual Risk of Malignancy in the Patients with Branch Duct Type Intraductal Papillary Mucinous Neoplasms of the Pancreas'. Together they form a unique fingerprint.

Cite this