Comparison of guidelines for intraductal papillary mucinous neoplasm: What is the next step beyond the current guidelines?

Abstract Management of intraductal papillary mucinous neoplasm is controversial, and several guidelines have aimed to establish an adequate strategy for surgical resection and surveillance. We compared various intraductal papillary mucinous neoplasm guidelines and considered new matters that are pivotal for improved treatment of intraductal papillary mucinous neoplasm. We identified and compared 11 published guidelines, three of which were major guidelines that mainly referred to the diagnosis and treatment of intraductal papillary mucinous neoplasm (International Association of Pancreatology 2012 guidelines, European Study Group on Cystic Tumours of the Pancreas 2013 guidelines, and American Gastroenterological Association 2015 guidelines). The main concerns of these three guidelines were indication for surgery and follow up of non‐resected lesions. Among the differences between the three guidelines, the period of surveillance recommended was the most controversial matter. Meanwhile, several nomograms have been proposed to improve the diagnosis of intraductal papillary mucinous neoplasm from the level of experts' experiences to that of rational systems. We discuss the adequate strategy of surveillance for intraductal papillary mucinous neoplasm with and without pancreatectomy and nomograms aiming to predict the risk of malignancy in patients with intraductal papillary mucinous neoplasm.

may develop distinct pancreatic ductal adenocarcinoma (PDAC) synchronously or metachronously. The natural history and high incidence of BD-IPMN make its surveillance controversial. Hence, several guidelines have been developed with an aim to establish an adequate strategy for surgical resection and surveillance of IPMN. We identified 11 available guidelines and further compared three major guidelines of IPMN published by the International Association of Pancreatology in 2012 (IAP2012), 11 European Study Group on Cystic Tumours of the Pancreas in 2013 (EURO), 12 and American Gastroenterological Association in 2015 (AGA). 13 We also herein present a discussion of new topics that are pivotal for the next step in improving the surveillance and treatment of IPMN.

| COMPARISON OF CURRENT
GUIDELINES FOR IPMN Table 1 shows 11 published guidelines concerning IPMN. Although they include other pancreatic lesions such as cystic neoplasms or pancreatic intraepithelial neoplasia, most of them focus mainly on management of IPMN. Among them, three guidelines deal with pathological issues (#1, #10, #11). [14][15][16] An illustrated consensus (#1) proposed a pathological definition of IPMN for differentiation from pancreatic intraepithelial neoplasia. 16  in differentiating IPMN from other cystic pancreatic lesions. 18 In 2006, international consensus guidelines (the IAP2006) (#3) were the first comprehensive guidelines referring to the diagnosis, indications for resection, and surveillance of IPMN. 19 The Society for Surgery of the Alimentary Tract Patient Care Guidelines (#4) provided general information on the categories, symptoms, diagnosis, and treatment of cystic neoplasms of the pancreas; however, these guidelines were unable to indicate either definite criteria for surgical resection or a surveillance strategy. 20 The consensus statements of the International Cancer of the Pancreas Screening consortium summit (#6) proposed that IPMN was a potential target for early detection and treatment in individuals at high risk for pancreatic cancer. 21 Italian consensus guidelines (#8) mainly focused on the diagnostic and follow-up strategies of pancreatic cystic neoplasms. 22 The remaining three guidelines (#5, #7, #9) are the current comprehensive guidelines citing diagnostic work-up, indications for surgery, surveillance after surgery, and surveillance of non-resected IPMN. [11][12][13] In this section, we compare these three guidelines ( Table 2). The IAP2012 (#5) and EURO (#7) guidelines are expert consensuses, whereas the AGA guidelines (#9) were established by an evidence-based approach using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework and the PICO (Patient problem or population, Intervention, Comparison and Outcome) format. However, the AGA guidelines noted that all evidence concerning the management of pancreatic cystic neoplasms was of very low quality because most of the available data were from retrospective case series.
Comparison of these three guidelines shows that the main concerns are indications for surgery and follow up of non-resected lesions.
Because IPMN varies from low-grade dysplasia to invasive cancer, it is important to establish a reliable indication that can predict malignant lesions. In the IAP2012 guidelines, high-risk stigmata (obstructive jaundice, an enhanced nodule, and a main pancreatic duct (MPD) of  Another point is surveillance after IPMN resection. For invasive IPMN, the IAP2012 and EURO guidelines recommend the same surveillance as carried out for pancreatic cancer. In contrast, the AGA guidelines suggest MRI surveillance every 2 years even after invasive IPMN. For noninvasive IPMN, the IAP2012 guidelines recommend repeat examinations at 2 and 5 years for new recurrences after resection, whereas they also suggest a 6-month interval, which is appropriate for surveillance considering the risk of PDAC development. The EURO guidelines recommend annual follow up for noninvasive IPMN. The AGA guidelines suggest MRI surveillance every 2 years after resection of high-grade IPMN, but they do not recommend routine surveillance after resection of pancreatic cysts without high-grade dysplasia or invasive malignancy. This matter will also be discussed later.

| PRED ICTION OF MALIGN ANT IPMN BY NOMOGRAM
Besides expert opinion-based guidelines, several efforts have been made to establish a nomogram as a more rational system with which to predict malignant IPMN, as listed in Table 3 8,25,26 and older age is also a predictor in two nomograms. 8,26 Gemenetzis et al. 29 included both MD-IPMN and BD-IPMN in a nomogram, and one of its factors predicting invasive cancer is the neutrophil-to-lymphocyte ratio, which has been reported to be a predictor of invasive cancer and poor prognosis in patients with various types of tumors. [30][31][32][33] These nomograms were validated by two methods: internal validation and external validation (Table 3). External evaluation of the nomogram is recommended because of objectivity and repeatability. 34 Three studies assessed their nomograms by external validation, 8,26,28 and one study used internal validation, 25 and one used no validation. 29 The concordance index and area under the curve derived from validation are used to estimate validity of the nomogram. External validation of the nomogram as established by Attiyeh  48 These data indicate that long-term surveillance (as long as the patient is fit for surgery) is required to detect remnant pancreatic lesions.
Hirono et al. 41 reported that a candidate risk factor for these remnant pancreatic lesions was dysplasia at the pancreatic cut margin, which included not only malignant lesions but also low-grade lesions. Pea et al. 51   In summary, it might be important to carry out long-term surveillance at short intervals for more than 5 years as long as the patient is fit for surgery. A common protocol for IPMN with and without resection is alternate CT and MRCP (EUS) twice a year in Japan; 60 however, its ability to improve overall survival and its cost-effectiveness should be evaluated. Further investigation using a prospective protocol with a large number of patients is needed to clarify the precise incidence of concomitant PDAC distinct from IPMN, to establish the optimal interval and period of surveillance, and to determine the most reliable risk factors for concomitant PDAC.

DISCLOSURE
Conflict of Interest: Authors declare no conflicts of interest for this article.