Introduction Endoscopic ultrasound (EUS) examination and fine needle aspiration (FNA)of the pancreas is now a routine method of obtaining diagnositic tissue for suspected pancreatic mass lesions, however if a biliary self expanding metal stent (SEMS) is in situ it can cast an acoustic shadow and cause local inflammation making EUS visualisation and FNA interpretation more challenging. Despite this it is common for a SEMS to be in place at the time of EUS due to the early treatment required for obstructive jaundice. Conflicting results have been published regarding the success of FNA in the presence of a SEMS, our aim was to analyse the accuracy of pancreatic FNA with or without a SEMS in an hepatobiliary-pancreatic cancer network in the United Kingdom.

Methods A retrospective analysis of all FNAs performed for solid pancreatic lesions in a single centre by 2 endosonographers from August 2010 to December 2012. Patents were identified if they had a biliary SEMS or plastic stent in situ. All FNAs were performed with 22 gauge FNA needle with no on site pathologist and direct expulsion of material into cyto-rich red medium. A standard 3 passes were performed in all lesions. Patient were excluded if they did not have a gold standard comparison for FNA results which was defined as surgical specimen comparison, death resulting from disease or clinical follow up and imaging for a minimum of 6 months.

Results 831 EUS procedure were performed during the study period of which 129 had FNA of a solid pancreatic lesions. 23 patents had a SEMS in situ at the time of EUS FNA and 5 had a plastic stent. The accuracy of pancreatic FNA with a SEMS was 65% (15/23)(95%CI: 44–81%) a plastic stent was 80%(4/5)(95%CI: 36% to 98%) and with no stent was 84.3% (86/102)(95%CI:76–90%) and overall was 82.2% (106/129)(95%CI: 75–88%). The FNA accuracy for the presence of a metallic stent was significantly lower χ² = 4.4 (p = 0.036). All 8 patients with a SEMS had a false negative result of which 5 were felt well enough to undergo a further procedure for consideration of chemotherapy which gave an accuracy of 80%(4/5). All patients were happy to undergo a repeat biopsy.

Conclusion The accuracy for FNA of solid pancreatic lesions in the presence of a SEMS can be significantly lower than without and should be taken into account when consenting patients and planning treatment. However it should not delay the insertion of a SEMS as if definitive cytology is required a repeat FNA is a feasible option with an acceptable accuracy.

Disclosure of Interest None Declared