Multidetector-row computed tomography (CT) of blunt pancreatic injuries: Can contrast-enhanced multiphasic CT detect pancreatic duct injuries?

作者：Wong Yon Cheong^*; Wang Li Jen; Fang Jen Feng; Lin Being Chuan; Ng Chip Jin; Chen Ray Jade

来源：Journal of Trauma-Injury Infection and Critical Care, 2008, 64(3): 666-672.

DOI：10.1097/TA.0b013e31802c5ba0

• blunt abdominal injuries
• pancreatic duct injuries
• multiphasic CT
• multidetector CT

Background: We examined patients of blunt trauma with contrast-enhanced multiphasic computed tomography (CT) and determined if it could detect pancreatic duct injuries.
Methods: During a 17-month period, 95 patients of blunt abdominal trauma underwent multiphasic CT examinations. The CT grading scales of pancreatic injuries at parenchymal phase, portal venous phase, and equilibrium phase were recorded and compared with surgery, endoscopic retrograde cholangiopancreatography, or discharged diagnosis. The diagnostic values of multiphasic CT and interobserver agreements at different phases were computed.
Results: Of the 95 patients, nine (9.5%) had pancreatic injuries (six with main duct injuries, three without main duct injuries). The interobserver agreement presented in kappa values between two radiologists regarding the integrity or disruption of the main duct were good at parenchymal phase (K = 0.73), portal venous phase (K = 0.64), and equilibrium phase (K = 0.68). The overall accuracies of multiphasic CT in detecting main duct injuries were 97.9% (parenchymal phase), 100.0% (portal venous phase), and 96.8% (equilibrium phase), respectively. The sensitivity (50.0%) and negative predictive value (96.7%) of equilibrium phase CT were the lowest among the three phases of CT scans.
Conclusion: The portal venous phase CT was the most accurate scan to detect pancreatic duct injuries. However, equilibrium phase CT might underestimate major pancreatic injuries. Multiphasic CT shows early promise in this clinical application and further multi-institutional studies to verify its accuracy and reveal the optimal CT methodology are needed.

• 出版日期2008-3
• 单位长春大学; 中国医科大学