PMO-098 Diagnostic yield of ERCP and EUS FNA in a tertiary pancreatic cancer centre

Introduction Patients with suspected pancreatic malignancy on cross sectional imaging are often referred for endoscopic investigations with a view to obtaining a definitive histological diagnosis. We aimed to assess the diagnostic yield of brushings and biopsies taken at endoscopic retrograde cholangiopancreatography (ERCP) and endoscopic ultrasound with fine needle aspiration (EUS FNA) in patients with pancreatic malignancy proven by operative histology. Methods A retrospective audit of 125 patients undergoing surgery for pancreatic malignancy at the Royal Liverpool University Hospital (RLUH) from January 2009 to December 2011 was carried out. Of these patients, 35 underwent investigation at RLUH and are included in this analysis, two of these required two investigations. Fifty patients had ERCP at an external Trust and 40 patients went straight to surgery without endoscopic intervention. Data were extracted from the pancreatic surgery database and electronic patient records for demographics and histology reports taken at the time of endoscopy and surgery. Results Overall 123 patients had operative histology confirming pancreatic adenocarcinoma, two patients had neuroendocrine tumour (NET) of the pancreas. Of the 35 patients undergoing investigation at RLUH, 34 had pancreatic adenocarcinoma and one had NET of the pancreas. Fifteen of 29 (52%) patients had brushings confirmatory of malignancy at ERCP, 9 of 29 (31%) had no malignant cells seen on brushings at ERCP, 3 of 29 (10%) had equivocal results suggestive but not diagnostic of malignancy. Brush cytology was not obtained in two patients, one patient suffered a perforation at ERCP requiring emergency surgery and one patient had failed cannulation of the CBD. Five of eight patients had an EUS FNA confirmatory of malignancy, two of eight had no malignant cells seen and 1 had equivocal results. In our cohort the sensitivity of ERCP alone is 56% (95% CI 36% to 74%) the combined sensitivity for ERCP and EUS 57% (95% CI 40% to 73%). Conclusion The sensitivity of brush cytology at ERCP for has previously been reported to be between 50% and 65% 1–3 for pancreatic malignancies. This is comparable to our findings in this cohort of patients with proven pancreatic malignancy. Positive histology from ERCP to EUS FNA can be confirmatory of pancreatic malignancy, but caution should be used when interpreting negative histology results given the intermediate sensitivity of these investigations. Competing interests None declared. References 1. Mansfield JC , Griffin SM, Wadehra V, et al. A prospective evaluation of cytology from biliary strictures. Gut 1997; 40 :671–7. 2. Macken E , et al. Brush cytology of ductal strictures during ERCP. Acta Gastroenterol Belg 2000; 63 :254–9. 3. Temino Lopez-Jurado R , et al. Diagnostic yeild of brush cytology for biliary stenosis during ERCP. Rev Esp Enferm Digest 2009; 101 :385–94.