A273 ENDOSCOPIC RETROGRADE CHOLANGIOPANCREATOGRAPHY IN A TERTIARY CANADIAN HOSPITAL: OPTIMIZING TISSUE SAMPLING TECHNIQUES

S A Taylor, A W Bak, T Lin, F Jowhari, M Savard, B Parker, J Dick, R Perini
2020 Journal of the Canadian Association of Gastroenterology  
Background Pancreatic and biliary cancer can be difficult to definitively diagnose, and tissue diagnosis defines clinical management in both early and late disease. The majority of patients with pancreaticobiliary malignancy present with late disease and only 30% have resectable tumours, contributing to poor prognosis and outcome. Endoscopic retrograde cholangiopancreatography (ERCP) is a mainstay for diagnosing and treating conditions of the bile and pancreatic duct. Aims This study aimed to
more » ... is study aimed to provide data on the diagnostic performance of ERCP utilizing conventional methods for tissue acquisition particularly in the setting of an indeterminate stricture. In particular, this was a quality assurance review to determine our centre's baseline performance as a reference so new methods of tissue acquisition such as SpyglassTM can be benchmarked against. Methods In a tertiary care hospital, 3723 ERCP procedures were performed between 2013 and 2017. Demographic and clinical information was retrospectively collected. Demographic and clinical information was collected. Three ERCP techniques of tissue sampling were analyzed: brushing alone (BiA), biopsy alone (BrA), and brushing and biopsy dual-modality approach (BrBi). The diagnostic performances (sensitivity, specificity, and accuracy) were calculated with data from post-sampling clinical reports, surgical pathological reports, and diagnostic imaging reports. Results 222 patients (285 ERCP procedure) met the study criteria and received fluoroscopy-guided ERCP sampling with cytology brushing and/or tissue biopsies. 36 (13%) patients had BiA, 85 (29%) patients had BrA, and 164 (58%) patients had BrBi. The sensitivities, specificities, and accuracies of the three sampling techniques respectively were: 56%, 93%, and 85% for BiA; 73%, 96%, and 83% for BrA; and 79%, 94%, and 86% for BrBi. Conclusions Our retrospective analysis of fluoroscopic-guided tissue acquisition shows accuracy as high as current published data on direct visualization cholangiosopy guided techniques. A prospective study looking not only at clinical yield but a cost analysis should be put in place prior to a commitment to purchase new equipment such as the SpyglassTM. Funding Agencies None
doi:10.1093/jcag/gwz047.272 fatcat:a4ygvzagfbd6dbdiwayzimsszi