The detection rate of Pancreatic cystic neoplasms (PCNs) as incidental finding has shown a continuous increase in the last decade, in particular in the elderly population. However, this is to be accounted more to the extensive use of cross-sectional imaging than to an effective increase of incidence in the general population [1]. Among the cystic lesions of the pancreas, intraductal papillary mucinous neoplasia (IPMN) is one of the most frequent [2]. IPMNs are characterized by intraductal proliferation of neoplastic mucinous cells with consequent cystic dilation of the pancreatic ducts and formation of macroscopic masses [2]. IPMN are classified in main duct type (MD-IPMN), branch duct type (BD-IPMN), and mixed type (MD + BD) taking into account the involved duct ( Fig. 1 ) [3]. 

IPMNs have a well-known potential of malignant transformation and are associated with an increased risk of pancreatic ductal adenocarcinoma (PDAC) in different sites within the gland [2, 3]. The probability of malignant transformation is relatively low (1.4% - 6.9% per year), and increases in elderly people [4-7]. The variability in malignant transformation requires an accurate stratification of at-risk patients, but the management of these patients is still under debate [8,9]. 

Several guidelines have been developed for stratification and management of patients with IPMN: the more widespread are the International Association of Pancreatology 2012 Guidelines v2017 (the Fukuoka Guidelines) [3], European Study Group on Cystic Tumors of the Pancreas 2013 Guidelines v2018 [10] and the American Gastroenterological Association (AGA) 2015 Guidelines [11]. All these guidelines have common key features and are widely validated. In particular, the Fukuoka Guidelines define the “worrisome features” (WF) and “high-risk stigmata” (HRS) to be assessed in CT or MRI studies for adequate stratification and management of patients with IPMN [3,12]. As an example, the presence of a WF is an indication for endoscopic US (EUS) evaluation and fine needle aspiration (FNA) while the presence of HRS is an indication for surgical resection [3,11]. 

Good reproducibility and inter-reader agreement (IRA) provide by standardized evaluation of imaging findings is desirable in terms of stratification, clinical management, and outcomes. Fukuoka and AGA guidelines demonstrated similar levels of accuracy in the detection of malignant transformation [13]. However, the Fukuoka Guidelines propose CT or MR as first-line evaluation, while other guidelines delegate this role to EUS; therefore, Fukuoka guidelines may have the theoretical potential of good reproducibility. 

The aim of this study is to retrospectively evaluate IRA of observers with different experience for pancreatic IPMN by using the Fukuoka guidelines.