None of these patients had new pain/discomfort or worsening of the baseline pain/discomfort at 24 hours after the procedure. None had procedure-induced pancreatitis. There were no other adverse events related to the procedure. The cytological diagnosis with the cell block method by H&E staining was positive (class IV or V) in 11 (Figure 3 and Figure 4) and negative (classes I, II, and III) in 33 (Table 1). Surgery was performed in 11 patients whose findings were

positive by cell block cytology (Fig. 5). Six patients with negative cytology results also underwent surgery AG-14699 because of mural nodules larger than 5 mm at first diagnosis in 4 patients and at progressive enlargement of more than 5 mm of the main and branch pancreatic ducts and mural nodules during follow-up on CT and EUS in the other 2 patients (Table 1). Histological analysis of the resected specimen revealed adenoma in 5 patients, in situ carcinoma in 8, and invasive carcinoma in 4 (Table 1). In the other 27 patients, the results did not indicate surgery and the patients were followed for more than 12 Vorinostat chemical structure months (range 13 to 50 months). They were regarded as having benign IPMNs because they showed no changes on CT or MRI imaging, including the diameter of the main and ectatic pancreatic ducts and the size of the mural nodule during

follow-up. Consequently, 73% (32/44) of the patients Interleukin-2 receptor were regarded as having nonmalignant IPMNs, and 27% (12/44) as having malignant IPMNs. There were no false-positive results and only 1 false-negative result. The sensitivity, specificity, and positive and negative predictive values of the cell block method for discriminating branch-duct type benign IPMNs from malignant ones were 92%, 100%, 100%, and 97%, respectively (Table 2). As for the immunohistochemical staining of mucin proteins, the cytological and histological results of MUCs 1, 2, 5AC, and 6 were in agreement in 88% (15/17), 94% (16/17), 88% (15/17), and 100% (17/17) of the

cases, respectively (Figure 3 and Figure 4; Table 3). At present, differentiation of benign and malignant IPMNs is still challenging. Although the International Consensus Guidelines are helpful regarding the management of IPMNs,18 the disadvantage of using these guidelines is the risk of overtreating patients. For example, only 15% of 61 patients with branch-duct type IPMNs who underwent resection had cancer according to a study on 147 patients by Pelaez-Luna et al.19 In our study, we demonstrated the usefulness of pancreatic duct lavage cytology with the cell block method for differentiating between benign and malignant branch-duct type IPMNs in patients having mural nodules.