The Sensitivity of Cross-Sectional Imaging for Identifying Periampullary Diverticula

Abstract

Introduction: Periampullary diverticula can make ERCP challenging due to unusual papilla position and distorted anatomy. Pre-ERCP CT or MRI, may help define anatomy, localize pathology, and identify periampullary diverticula. It is unclear whether technological advancements have led to improved radiographic sensitivity of periampullary diverticula or if adequate documentation is performed. Our aim was to determine the sensitivity of imaging in identifying duodenal diverticula as seen during ERCP. Methods: This is a retrospective study of 106 patients noted to have periampullary diverticula during ERCP between June 2014 and November 2017 at a single, large-volume academic center. The presence of duodenal diverticula noted on CT or MRI prior to ERCP was recorded. Age, gender, and diverticulum characteristics including number, size, and ampullary involvement were recorded. Successful cannulation was noted. The association between diverticula detection on imaging with diverticulum size, ampullary involvement, and biliary dilatation was determined using a Fisher's exact test. Results: A total 3289 ERCPs were performed in the study period, in which 106 patients were noted to have periampullary diverticula on ERCP. Of these 106 patients, 31 had CT, 52 had MRI, and 23 had both. Detection rate of diverticula on CT or MRI was 25.7% (28/106). Most patients had a single diverticulum (n=96, 88.1%) and ampullary involvement (n=76, 69.7%). Biliary ductal dilatation was noted on imaging in 68.7% (57/83). For the 23 individuals who underwent both CT and MRI, there was a 17.4% discordance rate. Diverticula were detected on CT or MRI in 18 of 51 (35.2%) patients with large diverticula, and 9 of 54 (16.7%) small/medium-sized diverticula (p=0.044). There was also a significant difference in detection of diverticula between patients with versus without a dilated bile duct (28.1% vs. 23.1%, p=0.0001). No significant difference was seen in detection of diverticula with or without ampullary involvement (28.9% vs. 20.0%, p=0.467).1154_A Figure 1. Patient characteristics, diverticulum qualities, procedural details, and indication for ERCPConclusion: The sensitivity of cross-sectional imaging for detecting duodenal diverticula among patients undergoing ERCP remains low despite advancements in imaging. There was a statistically significant difference in sensitivity of diverticulum detection based upon diverticulum size and biliary dilatation, but not ampullary involvement. This study highlights the potential to improve detection of periampullary diverticula by radiologists, which may impact pre-procedural planning and outcomes.1154_B Figure 2. Differences in diverticula detection on cross-sectional imaging by imaging modality, diverticulum size, ampullary involvement, or bile duct dilatation