Computed tomography classification of endoscopic retrograde cholangiopancreatography‐related perforation

Abstract

Endoscopic retrograde cholangiopancreatography (ERCP)‐related perforation leads to high morbidity and mortality. The Stapfer classification divides patients with different perforation locations and suggests management accordingly. The classification may be unknown if perforation is not detected during endoscopy. We classified patients with ERCP‐related perforation (ERP) through computed tomography (CT) and observed the clinical outcomes with varyingly invasive management. Fifty‐two cases of ERP between July 2009 and December 2017 were retrospectively reviewed. Of them, 41 who underwent CT for ERCP were included. According to their CT findings, we divided patients into air‐alone (n = 16), air‐fluid (n = 18), and fluid‐alone (n = 7) groups. Perforation severity was graded using the Clavien‐Dindo classification for surgical complications. Demographic data and clinical outcomes among different groups were analyzed. Fifteen patients (37%) had an unknown Stapfer classification. More than half of the patients in the air‐fluid group had a Clavien‐Dindo complication grade of >3. Four patients underwent surgical repair; all of them were from the air‐fluid group. All patients in the air‐ and fluid‐alone groups underwent medical treatment without need for subsequent salvage surgery. The air‐fluid group had the longest mean hospital stay (25.1 ± 21.9 days) and the exclusive two mortality cases in this study. Patients with ERCP can be divided into groups with different outcomes according to the presence of air or fluid on CT images. Because patients with both air and fluid have the worst clinical outcome, they may require more aggressive treatment than patients with either air or fluid alone.

Keywords

• air
• computed tomography
• endoscopic retrograde cholangiopancreatography
• fluid
• perforation