Eur Radiol (2002) 12:979–993 DOI 10.1007/s00330-002-1427-x
R.F. Dondelinger G. Trotteur B. Ghaye D. Szapiro
Published online: 23 March 2002 © Springer-Verlag 2002 Categorial Course ECR 2003
R.F. Dondelinger (✉) · G. Trotteur B. Ghaye · D. Szapiro Department of Medical Imaging, University Hospital Sart Tilman, Domaine Universitaire du Sart Tilman B35, 4000 Liege, Belgium e-mail:
[email protected] Tel.: +32-4-3667259 Fax: +32-4-3667224
EMERGENCY RADIOLOGY
Traumatic injuries: radiological hemostatic intervention at admission
Abstract Blunt trauma victims and selected patients with penetrating trauma are systematically investigated after resuscitation and hemodynamic stabilization with cross-sectional imaging. Computed tomography is a good predictor of the need for hemostatic arteriographic embolization, based on contrast medium extravasation observed on CT. In centers admitting polytrauma patients, the CT and angiography units should be installed together within the emergency environment. Trauma-dedicated interventional radiologists should be on call for optimal patient management. Posttraumatic retroperitoneal and pelvic bleeding is a primary indication for angiographic hemostasis, together with orthopedic fixation of pelvic bone fractures.
Introduction Currently, suicide and trauma are the leading causes of death in the male population under 25 and 40 years of age, respectively, and are responsible for the loss of more years of life than cancer and cardiovascular diseases together [1]. Traumatic injuries of the central nervous system, heart, and great vessels are responsible for the deaths that occur at the site of accident, whereas visceral hemorrhage is the principal cause of mortality during the first 4 h following trauma. The vast majority of nonbleeding internal injuries are compatible with survival; therefore, urgent patient admission (“load and go”), rapid resuscitation and hemodynamic stabilization are the prerequisites for setting management priorities, followed when possible by aggressive and reliable investigation of
Angiography should be carried out rapidly, before the patient decompensates for considerable blood loss. In patients with visceral bleeding, arterial embolization can obviate primary surgery or potentializes surgical intervention and contributes to changing hierarchy of injuries to be treated surgically. Failure to achieve primary hemostasis may occur according to the type of specific organ injury and coagulation and metabolic parameters of the patient. Postembolization complications are few and are usually non-life-threatening and rarely carry definitive sequelae. Keywords Injuries · Computed tomography · Angiography · Hemostatic embolization procedures
all posttraumatic injuries for the selection of appropriate treatment. Urgent intervention is the cornerstone of hemostasis. Arteriographic embolization has been shown for several decades to represent a useful adjunct in the treatment of persistent or recurrent postoperative bleeding, and to be a reasonable primary alternative to surgery in retroperitoneal or pelvic bleeding, or in selected patients with abdominal visceral hemorrhage.
Classification of trauma patients History taking of the mechanism of trauma is helpful in anticipating the presence of potential posttraumatic injuries and in sensing their severity. Despite the availability
980
of powerful imaging, the gross clinical findings always take precedence in the sequence of management and imaging. Trauma victims can be classified into three categories for management orientation: 1. Patients in class I, who remain hemodynamically unstable despite resuscitation, are directed to the operating room. “Focused Assessment with Sonography for Trauma” (FAST) or a standardized six-point trauma ultrasonography examination was introduced as a means of rapid evaluation of the abdomen (and pericardium) in the shock room, and has replaced diagnostic peritoneal lavage, to decide on immediate laparotomy or thoracotomy, without additional imaging [2, 3, 4, 5]. 2. Patients in class II are marginally stable, and will probably require surgery; however, when ongoing retroperitoneal hemorrhage is suspected, these patients should undergo angiography [6]. Patients with pelvic bone fractures primarily undergo external orthopedic fixation, followed by angiographic evaluation, when pelvic bleeding persists. Presently, many of the marginally stable patients benefit from CT, preceding orthopedic fixation, and therapeutic pelvic or retroperitoneal angiography. Marginally stable patients with suspected thoracic or abdominal visceral bleeding can also undergo CT, if the local logistics allow performance of the examination without delay in the resuscitation environment. 3. Patients in class III are hemodynamically stable, and are systematically evaluated with CT. Some patients with clinical suspicion of minimal single-organ injury may undergo US or CT without the need for a formal hospital admission. In summary, according to this classification, based on the hemodynamic status, patients either undergo immediate surgery, or hemostatic angiography, alone or in combination with surgery, or they are included in a trial of expectant conservative management.
Trauma patient triage with CT Computed tomography largely contributes to management orientation of trauma victims. Spiral CT has been recognized to be the modality of choice for investigation of blunt trauma, allowing for a rapid screening of injuries of the brain, face, thorax, abdomen, spine, and pelvis. Significance of hemoperitoneum for patient triage Both CT and US are able to detect with confidence hemoperitoneum and have largely superseded diagnostic
peritoneal lavage; however, the observation of hemoperitoneum is relatively irrelevant for management of stable patients. Peritoneal fluid confirms visceral bleeding, but is not a reliable indicator of need for hemostasis. In our experience, 1% of all adult patients with blunt abdominal trauma have hemoperitoneum without detectable abdominal organ injury and 4.3% have no intraperitoneal injury [7]. In children, 11% were reported to have “unexplained” isolated peritoneal fluid [8]. In these patients, hemoperitoneum may be caused either by minimal mesenteric bleeding, or by extension of retroperitoneal hematoma into the peritoneal cavity through a ruptured parietal peritoneal leaflet [9, 10]. On the other hand, we and others have observed parenchymal hepatosplenic injuries of moderate severity grade and without hemoperitoneum in 11–25% of the cases, when CT was urgently performed after trauma [7, 11, 12, 13]. Significance of organ injury demonstration and injury severity grading There is general agreement, that cross-sectional imaging evidences hemoperitoneum more easily than specific abdominal organ injury. A significant false negative rate of 20–30% might be expected in the diagnosis of parenchymal injuries when using FAST. In hemodynamically stable patients, FAST was shown unreliable in comparison with CT, with a sensitivity of only 30% [14]. Furthermore, in polytrauma patients, US is not applicable to brain, lung and bone injury. Early CT reports on the first 200–500 patients that were investigated for abdominal or pelvic trauma showed that there were only a few false-positive or falsenegative interpretations of abdominal examinations [15, 16]. Most papers published thereafter credited CT with overall sensitivities, higher than 90%, and specificities close to 100%, in the diagnosis of abdominal parenchymal injuries in adults and in the pediatric age group. More differentiated injury-specific sensitivities were published for the pancreas (0%), gastrointestinal tract (42%), and urinary bladder (50%) [17]. Some publications also reported lower sensitivities, of 85, 84, and 67% for renal, hepatic, and splenic injuries [18]; however, the accepted statement is that only a few percent of errors are observed in thoracoabdominal CT, and most missed injuries are generally not responsible of posttraumatic death, neither do they carry a life-threatening posttraumatic complication [17]. The amount of hemoperitoneum correlates with the need for laparotomy, but the demonstration of high-grade parenchymal injury is not a good predictor for intervention. The amount of hemoperitoneum correlates with the severity of splenic injury, but not with hepatic injury in our experience [7]. High-grade (III–IV) splenic and hepatic injury may require more often hemostatic interven-
981
tion than lower-grade injury, but low-grade liver trauma (grades I and II) responds to expectant non-operative treatment more consistently than splenic injury. Demonstration of ongoing hemorrhage It is well established that overall 50–70% of all liver and splenic injuries have stopped bleeding at time of operation and can be treated conservatively, without surgery [19, 20, 21, 22, 23]; therefore, imaging should be able to disclose “active bleeding,” which is expected to serve best as a triage of patients for urgent hemostatic intervention or expectant conservative treatment. As CT is able to detect minimal density gradients, it has been hypothesized that small amounts of extravasated contrast medium could be recognized, and that CT could give equivalent or superior results to angiography in the diagnosis of active hemorrhage. In fact, ongoing bleeding (25–370 HU, mean 132 HU) can be differentiated from clotted blood (40–70 HU) by measurement of CT attenuation values (p