Introduction The risks of routine preoperative biliary drainage in pancreatic cancer patients without cholangitis and bilirubin <250μmol/L are established. However, access to early resection without prior biliary stenting is challenging in most UK hepatobiliary (HPB) centres. We evaluated adherence to the European Society of Gastrointestinal Endoscopy guidelines (ESGE 2012) and reviewed clinical and stent outcomes in patients referred through the East Midlands Cancer Network.

Methods Patients with a diagnosis of pancreatic cancer during 2014 were identified and cross-referenced with our HPB surgical database. We examined patient demographics, diagnostic modality, tumour site and type, whether pre-operative drainage was performed with endoscopic retrograde cholangio-pancreatography (ERCP) or percutaneous trans-hepatic cholangiography (PTC) and stent choice. Clinical outcomes included ERCP and surgical technical success as well as re-intervention rates.

Results 135 pancreatic cancer cases (70 men; mean age 69±8 years, 65 women; mean age 72±7 years) were identified. Pathological diagnosis from EUS-FNA (n = 42), EUS-FNB (n = 34) or brush cytology (n = 24) was adenocarcinoma (n = 95), adenosqamous carcinoma (n = 2), mucinous adenocarcinoma (n = 2), neuroendocrine tumour (NETs; n = 9), squamous carcinoma (n = 2) and 25 radiological diagnosis alone. 50/135 (37%) patients underwent ERCP with technical success in 29/50 (58%): 14 plastic stents; 10 fr (n = 10), 7 fr (n = 4); and 15 self-expanding metal stents (SEMS); 10 mm fc-SEMS; 4 cm (n = 2), 6 cm (n = 2), 8 cm (n = 2); 10 mm uc-SEMS; 4 cm (n = 3), 6 cm (n = 5) were inserted. ERCP failure was due to either unsuccessful CBD cannulation and stenting (15/21; 71%) or duodenal obstruction (6/21; 29%) and 16/21 (76%) of these had PTC, were resected 2/21 (9%) or palliated 3/21 (14%). Re-intervention following ERCP stenting was required in 8/29 (28%): plastic stents 6/8 (75%) obstruction/cholangitis, SEMS 2/8 (25%) distal migration. 10/15 (67%) with SEMS inserted died <4 months after ERCP.

27/135 (20%) cases underwent surgery: complete resection 10/27 (R₀; 37%), incomplete resection 7/27 (R₁; 26%), palliative bypass 7/27 (26%) after a mean delay of 26 d (adenocarcinoma) and 90 d (NETs). 3/27 (11%) developed disease recurrence. In patients with biliary obstruction (9/27 who underwent surgery), 5/9 (55%) had surgical resection without attempted pre-operative drainage, with mean bilirubin 144μmol/L.

Conclusion Following diagnosis of pancreatic cancer, few patients had surgical resection without prior drainage and overall <10% achieved cure. Success in ERCP biliary drainage should be improved and EUS-guided biliary access explored. Although patency rates are higher with SEMS, a high proportion died <4 m, suggesting better case selection for SEMS is required.

Disclosure of Interest None Declared