Management of Pancreatic Cystic Neoplasms: Decision-Making with Limited Information

available imaging studies, often done elsewhere. The preoperative diagnosis was incorrect in 32% of the cases in the immediate surgery group and 48% of the cases in the observation group. These data suggest that some patients who should be resected are likely being followed (e.g., patients with microscopic main duct involvement not apparent on imaging or small MCNs) and some patients are subjected to surgery when they do not need it (e.g., those with non-neoplastic cysts). It is quite clear from this study and our experience that in cystic pancreatic neoplasms the preoperative diagnosis does not always correlate with the final pathologic diagnosis. In some cases it is nearly impossible to avoid resecting a large non-neoplastic cyst due to inconclusive preoperative tests or tests suggesting that it is a mucinous cyst. While it is sobering to note that we are making important decisions regarding the need for major pancreatic resection based on less than optimal data, a closer examination of the study data shows that the clinical impact of this may not be as severe. In some cases additional testing may have given a definitive preoperative diagnosis, but would not have changed the decision to resect (e.g., a solid pseudopapillary tumor). While frank main duct involvement is associated with a high prevalence of malignancy, there are no data on the prognostic implications of microscopic main duct involvement. Sendai criteria are based on P ancreatic cystic lesions have a broad differential diagnosis, from benign cysts (e.g. pseudocyst), potentially malignant cysts (intraductal papillary mucinous neoplasm, IPMN, and mucinous cystic neoplasm, MCN) to frankly malignant cysts (e.g., ductal adenocarcinoma, mucinous adenocarcinoma and islet cell cancers). Recently published consensus guidelines (a.k.a. Sendai criteria) recommend resection for all cysts which are presumed to be MCNs based on the fact that they are solitary, left-sided and cured by distal pancreatectomy. For presumed branch duct IPMN, resection is recommended if 6 1 of these features is present: (i) symptoms; (ii) cyst size 6 3 cm; (iii) main duct size 1 6 mm; (iv) intramural nodules, or (v) positive malignant cytology on cyst fluid [1] . In patients with presumed mucinous cysts, the Sendai criteria identify those with a higher likelihood of having a high-grade lesion (carcinoma in situ or invasive cancer). Application of Sendai criteria requires assessment of symptoms, findings on cross-sectional imaging and endoscopic ultrasonography (EUS) with cyst fluid aspiration. In this issue, Correa-Gallego et al. [2] correlated preand postoperative diagnosis in incidentally discovered cysts that were operated on immediately (n = 136) or after an initial period of observation (n = 23). Decision to operate was based on Sendai criteria applied to data from Published online: May 12, 2010