Yield of Pancreatic Juice Cytology (PJC) in the Diagnosis of Strictures of the Pancreas Duct

Abstract

Background: Despite advances in technology and methods, tissue confirmation of suspected malignant strictures (MS) of the pancreatic duct (PD) remain a challenge. Prior to chemotherapy or radiation therapy, tissue confirmation is required. Lack of access to strictures of the PD with near or total obstruction may preclude passage of a brush upstream to the stricture, making brush cytology impossible. In situations such as this, PJC may be the only method available for tissue confirmation. Methods: The ERCP database was searched for pts having PJC from 1994-2003. Pts having PJC for the diagnosis of mucin producing tumors (MPT) were excluded, since the suspicion of having a positive result in these patients would be low. Results: PJC was collected in 257 patients over 10 years. The final diagnosis was MPT in 200 patients, while a total of 57 patients had PJC collected for the diagnosis of a suspected MS of PD. The median age of 25 males was 57 years and for 32 females was 61 years. The stricture was located in the ampulla (1), head (30), genu (8), body (16), or tail (2). In 22 patients, MS was confirmed by one or more of the following methods: CBD brushing (2), EUS FNA (9), surgical resection (11), CT FNA (1), and PD stent cytology (1). Only 2 of these (9%) had a positive PJC. One of these revealed adenosquamous carcinoma (ASC) and the other MPT. The MPT was not confirmed on EUS FNA, but adenocarcinoma (AC) was found at surgical resection. A total of 12 pts with suspected benign strictures by PJC had negative tissue confirmation at EUS FNA or at surgical resection. PJC had a sensitivity of 9.1%, specificity of 100%, positive predictive value of 100%, and negative predictive value (NPV) of 37.5%. An additional 37 pts had negative PJC, without confirmation, giving a total of 49 negative PJC (86%). The 8 pts without a negative PJC had a variety of results. A single positive result for carcinoma was found and confirmed with distal pancreatectomy, pathology showed ASC. A single PJC revealed rare atypical cells, and PD brushing confirmed AC. Another showed fungal elements and AC was found in both CBD brushing and EUS FNA. A single sample revealed inflammation, with a negative CBD brush and rare atypia on PD brushing. AC was confirmed later with cytology from the PD stent. Three pts had evidence of MPT. One of these had AC confirmed on EUS FNA and at Whipple procedure. Summary: PJC has a very low sensitivity and NPV in patients with suspected MD of the PD. Conclusions: While PJC is a valuable tool in the diagnosis of MPT, it has a low yield in suspected MS of the PD.