[Contrast media extravasation in upper abdominal injuries: detection with spiral computerized tomography]

The possibility of detecting contrast agent extravasation (i.e., active hemorrhage) with dynamic conventional Computed Tomography (CT) in patients with abdominal trauma has already been reported in small series. We report our experience in the demonstration of contrast material extravasation using helical CT; we also investigate the diagnostic and clinical value of this finding.

January 1997 to July 1998, we examined 41 consecutive patients with upper abdominal trauma. Twelve patients (29%) had contrast material extravasation. The examinations were performed with a helical unit and volumetric acquisitions (thickness 8-10 mm, pitch 1, reconstruction interval 5-8 mm). The intravenous contrast medium (350 mgI/mL, 130-140 mL) was administered with rapid infusion (2-2.5 mL/s, 40-50 s acquisition delay from bolus starting) and using a power injector. We reviewed the CT studies and clinical records of these 12 patients. Contrast agent extravasation was considered present when this finding, not recognizable on plain scans, showed equal attenuation to or higher attenuation than the vessels within the same level. Moreover we assessed leak site, CT appearance, the direct visualization of the involved vessel, the evidence of other abdominal or extra-abdominal injuries, the CT signs of hypovolemic shock, clinical and surgical data. For comparison, we finally evaluated 50 examinations performed with a conventional CT scanner in subjects with abdominal trauma.

Active hemorrhage involved the abdominal wall in 1 case (intercostal artery), the solid organs in 4 (splenic in 2, hepatic in 1, of the middle hepatic vein in 1), the peritoneal cavity in 3 (splenic, midcolic, and gastroduodenal artery in 1 each), the retroperitoneum in 4 (renal pedicle in 2, renal parenchyma in 1, lumbar artery in 1). In all cases the site of contrast extravasation corresponded at surgery to the site of active bleeding. The pattern was localized in 10 cases and diffuse in 2. The involved vessel could be identified in 5 cases while in the other ones the origin could be inferred from the leakage site. Associated injuries of upper abdominal organs were seen in 11 of 12 patients and extra-abdominal trauma in 6. In 4 cases there were CT features of hypovolemia. One patient died during transport to the operating room and another after surgery, while all the others survived. Contrast extravasation was identified in 9 (18%) of the patients examined with a conventional CT unit.

Active contrast material extravasation can be recognized with conventional CT scanners, though it has been considered a rare finding. Helical CT seems to increase the detection rate and especially to boost the radiologist's confidence in this diagnosis. Though active bleeding is identified in severely-injured subjects requiring urgent intervention and may be associated with findings of hypovolemic shock, it should not be considered itself as a negative prognostic factor. Contrast extravasation is due to ongoing hemorrhage and its detection is critical for urgent treatment. Accurate anatomical location permits to choose surgical management or transcatheter embolization and thus decreases time consumption for precise bleeding site identification.