INDICATIONS

The indication for using interventional radiology techniques in the care of trauma patient depends on the technique being used and the medical condition of the trauma patient. It is not appropriate to transport a hemodynamically abnormal patient to the angiographic suite for a diagnostic arteriogram. The arteriogram should either be performed in the operating room or delayed until the patient is resuscitated. Alternatively, patients with severe pelvic trauma should undergo relatively prompt arteriography and possible embolization to try to achieve hemostasis. These patients may be hemodynamically abnormal but the interventional radiology techniques are part of their resuscitation. In other injuries, interventional radiology techniques may be used to provide definitive therapy for hemorrhage. The development of intravascular covered stents has allowed for the management of certain injuries, including arterial disruption and arteriovenous fistulas, to move from the operating room to the angiographic suite. Venous injuries including lacerations, thrombosis, and compression can be managed by the placement of uncovered intravascular stents. The decision to use interventional radiology techniques and the timing of these techniques requires that the trauma surgeon understands the indications and limitations of these techniques.

BLUNT CEREBROVASCULAR INJURY

Over the past decade, there has been an increased awareness of blunt cerebrovascular injury (BCVI), which includes carotid artery injuries (CAI) and vertebral artery injuries (VAI). Several studies have identified an incidence of 1.5%–2% for BCVI among blunt trauma victims. Historically, blunt carotid artery injuries had been diagnosed by onset of neurologic symptoms. The outcomes for this injury were poor. Blunt carotid injuries had an associated mortality rate of 31% and a stroke rate of 43%. Blunt carotid injuries have been shown to have worse outcomes compared to penetrating carotid injuries. Seventy-eight percent of patients with penetrating carotid injuries have been found to be independent with locomotion at the time of discharge compared to 37% of those with blunt carotid injuries. Blunt carotid injuries also have a worse outcome compared to the overall blunt trauma population. Fifty-five percent of blunt trauma patients are able to be fully independent at the time of discharge compared to 33% of patients with blunt carotid injury. VAI occurs in 0.53% of blunt trauma patients, and has an associated stroke rate of 25%. Screening protocols using four-vessel angiography have been used to successfully identify these injuries prior to the development of neurologic symptoms. The institution of early treatment with anticoagulation or antiplatelet therapy in VAI has been shown to decrease the stroke rate from 14% to 0%. When anticoagulation or antiplatelet therapy is used in the management of blunt carotid injuries, the stroke rate decreases from 60% to less than 10%. It is clear that there is significant morbidity and mortality when BCVI is missed. However, they occur relatively infrequently, such that it is not practical to screen all blunt trauma patients.

Several screening triggers have been suggested in the literature including cervical spine fracture, neurologic findings not explained by radiographic findings, Horner’s syndrome, LeFort II or III facial fractures, skull base fractures involving the foramen lacerum, and neck soft tissue injury. A good screening test should be relatively inexpensive, have a low morbidity rate, and a high sensitivity rate. It should find all the true positive results with some false positives and no false negatives. In a comparison of magnetic resonance angiography (MRA), computed tomographic angiography (CTA), and four-vessel cerebral angiography between 2000 and 2002, the sensitivity of MRA and CTA for BCVI was 47%–53%. These rates are too low for a test to be an effective screening modality. Four-vessel cerebral angiography has been identified as the gold standard for the diagnosis of BCVI. However, its cost and major complication rate of 1%–3% in large series make it a less than ideal screening test. The development of the multidetector CT scanner has increased the resolution of the CT scanner. Two recent studies demonstrated that CTA performed on multidetector CT scanners has dramatically improved ability to diagnose these injuries. A head-to-head comparison of CTA with multidetector CT scanners and four-vessel cerebral angiography has yet to be done. As a result, four-vessel cerebral angiography remains the gold standard for diagnosis and screening of these injuries. It will be important to continue to monitor improvements in CTA, MRA, and possibly even ultrasound technologies for less invasive, cheaper, and safer screening modalities for BCVI.

Blunt cerebrovascular injuries have been effectively managed by anticoagulation or antiplatelet therapy in patients with contraindications to anticoagulation. With the development of endovascular technologies including balloons, coils, and stents, there have been several single-institution, small series that have demonstrated good efficacy in the treatment of traumatic pseudoaneurysms of the carotid artery with good short-term follow-up. However, one large series with 46 patients over an 8-year period demonstrated a 21% complication rate and 45% occlusion rate for patients treated with endovascular stents. Patients who had received antithrombotic agents alone only had a 5% occlusion rate. This study concluded that antithrombotic therapy was the recommended therapy for blunt carotid injuries, and that the role of stents remains undefined. The indications for the use of endovascular techniques in the management of BCVI remain unclear and their long-term results are unproven. Clearly, further study of these techniques is warranted.

BRACHIOCEPHALIC TRAUMA

The use of diagnostic angiography in penetrating neck injuries is based on which zone of the neck is involved. Penetrating injuries to zone 1 of the neck are usually evaluated with angiography of the carotid and subclavian arteries, if the patient is hemodynamically normal. Penetrating injuries to zone 2 of the neck can be managed by unilateral or bilateral neck exploration. If the patient is hemodynamically normal, evaluation of the injury with four-vessel cerebral angiography, esophagoscopy or soluble-contrast esophagography, and bronchoscopy can be performed. Some authors have reported using a multidetector CT scan of the neck with CT angiography (CTA) to evaluate these injuries. At this time, CTA has not been directly compared with four-vessel cerebral angiography and future studies will determine whether CTA can replace four-vessel cerebral angiography. Penetrating injuries to zone 3 of the neck are usually evaluated with four-vessel cerebral angiography because of the difficulty in achieving adequate surgical exposure of either the internal carotid artery or the vertebral artery at this level. Endovascular techniques including coil embolization and endovascular stents have been used successfully to manage vascular injuries in this area of the neck.

Angioembolization has also been successfully used to control hemorrhage associated with blunt and penetrating facial injuries. Attempting to obtain hemostasis of these injuries operatively can be very difficult. As a result, arteriography with angioembolization has become the first-line treatment for many of these injuries.

There have been several case and short series reports regarding the use of endovascular stents to repair blunt and penetrating injuries to the subclavian, axillary, and brachial arteries. All of them have demonstrated excellent success in acutely managing these injuries and few complications in short-term follow-up. The long-term success of these devices remains unknown.

THORACIC INJURY

Blunt aortic injury (BAI) may be caused by either rapid horizontal or vertical deceleration. Rapid horizontal deceleration usually results in an injury to the aorta at the isthmus, just distal to the origin of the left subclavian artery. The injury may be a dissection or transection of the aorta which can lead to hemorrhage, pseudoaneurysm formation, or thrombosis. Seventy to 90% of patients with traumatic injury to the thoracic aorta die at the scene. An additional 30% of these remaining patients die in the hospital prior to undergoing definitive surgical treatment. Surgical repair has been associated with mortality rates between 5%–25% and spinal cord infarction rates as high as 13%. It remains important to promptly diagnose this injury so that the patient can be appropriately managed. BAI may be suggested on routine chest radiograph. Findings such as a widened superior mediastinum, an indistinct aortic knob, or widening of the paratracheal stripe are all suggestive of a BAI. However, the specificity of these findings is only about 5%–10%. A CT scan of the chest may be performed to further delineate the aorta. The absence of a mediastinal hematoma on a helical CT scan of the chest has been shown to have a negative predictive value of 97%–99%. The evolution of the multi-detector CT scanner over the last 5 years has allowed for the development of CT angiography (CTA), which appears to have high sensitivity and specificity according to preliminary reports. Some institutions have begun using CTA as their diagnostic test of choice for BAI. At this time, aortography remains the gold standard for diagnosis of BAI, but it will most likely be replaced by CTA if its sensitivity and specificity are as high as early studies suggest. Alternatively, patients with evidence of blunt aortic injury could undergo transesophageal echocardiography to evaluate the aorta. The decision to obtain aortography or transesophageal echocardiography depends on which of these modalities is more readily available to the trauma team and the preference of the surgeon performing the repair of the aorta.

The management of BAI has been changing over the last several years. If the patient is hemodynamically normal, the patient is admitted to the ICU and conservatively managed until the associated injuries are resolved. Conservative management includes the use of beta-blockers and vasodilators to prevent tachycardia and hypertension. Several studies have demonstrated a decreased rate of in-hospital mortality and paraplegia with delayed repair of BAI. Endovascular stents have been used to repair BAI in both the emergent and delayed settings. The reported cases have all demonstrated few acute complications and 30-day mortality rates below 15%. While these results are encouraging, long-term results are unknown. One long-term study on endovascular repair of aortic aneurysms demonstrated a 65% endoleak rate, 35% graft migration, and 78% graft deformation. In fairness, this study involved some of the first generation of endovascular stents, and hopefully, the results for newer generation of stents will be better. Figure 1 demonstrates a traumatic pseudoaneurysm of the descending aorta. Figure 2 demonstrates the completion aortogram after placement of an endovascular stent. The first commercially available endovascular stents for the thoracic aorta only became available in the United States in 2005. There is currently a prospective analysis of the use of these grafts in the management of BAI being conducted by the American Association for the Surgery of Trauma (AAST). The ultimate success of these devices will be based on their long-term performance. It is important to remember that the average age of a trauma patient is much lower than that of patients undergoing aortic aneurysm or dissection repair. These grafts may need to last 40–60 years when placed in a young trauma patient. These results are simply unknown at this time.

Figure 1 Traumatic pseudoaneurysm of the descending aorta.

Figure 2 Completion aortogram after placement of endovascular stent.

Angiography can be useful in the evaluation of victims of both penetrating and blunt trauma. Patients who have suffered transmediastinal penetrating injuries should undergo aortography as part of their evaluation. CTA using multidetector CT scanners is still being evaluated, but it will most likely replace conventional catheter aortography as the screening test of choice.

ABDOMINAL TRAUMA

The two most commonly injured abdominal organs in blunt trauma are the liver and spleen. If patients are hemodynamically abnormal and have evidence of abdominal injury, either by a positive diagnostic peritoneal lavage (DPL) or a positive focused abdominal sonogram for trauma (FAST), the patient is taken to the operating room for surgical exploration. If the victim of blunt trauma is hemodynamically normal, they are evaluated with a contrast-enhanced CT scan of the abdomen. If the patient is hemodynamically normal, has a Glasgow coma score (GCS) of 15, does not have a distracting injury, and is not impaired by alcohol or illicit drugs, the patient’s abdomen can be evaluated by clinical examination. The development of the CT scanner has allowed for the nonoperative diagnosis of injuries to the spleen and liver and the development of scoring systems for these injuries. During the late 1980s and early 1990s, several large studies demonstrated that hemodynamically normal patients with blunt injuries to the spleen and liver could be successfully managed nonoperatively with success rates of 70%–85%. This is now the routine practice for the management of blunt injury to the liver and spleen in hemodynamically normal patients.

At the same time that nonoperative management of blunt splenic injuries was developing, there were some case reports of using angioembolization to manage hemorrhage from blunt splenic injury. As a result, some groups of trauma surgeons began routinely performing arteriography on all blunt splenic injuries that qualified for nonoperative management. Any evidence of active hemorrhage or pseudoaneurysm formation was managed with angioembolization. This included proximal splenic artery embolization with coils or large gelfoam and more selective embolization of arterial branches with smaller coils or gelfoam. This strategy increased the success rate of nonoperative management of blunt splenic injuries to 93%–97%. Other groups looked for more selective criteria for using angioembolization in the nonoperative management of blunt splenic injuries. Several studies demonstrated that evidence ofactive contrast extravasation or pseudoaneurysm formation on contrast-enhanced CT scan was predictive of failure of nonoperative management. Angioembolization was used as an adjunct in patients with evidence of active contrast extravasation or pseudoaneurysm to achieve nonoperative management success rates similar to those achieved with the routine use of arteriography on all splenic injuries. As a result, most institutions have incorporated angioembolization into their nonoperative management algorithms for splenic injuries in a selective fashion, usually based on findings on CT scan. There is a low complication rate for angioembolization of the spleen, which includes total splenic infarction, splenic abscess formation and complications, related to the arterial access for the procedure. This complication rate is lower when selective angioembolization of branch vessels of the splenic artery is performed rather than occlusion of the main splenic artery. In addition, angioembolization is not technically feasible in every patient. However, despite these few limitations and complications, angioembolization has clearly established itself as a useful adjunct to the nonoperative management of blunt splenic injuries. It is now possible to manage 70%–75% of all blunt splenic injuries nonoperatively with success rates over 90%.

The role of angioembolization in the management of blunt hepatic trauma has developed over the last 25 years as well. Initially, angioembolization was reported as being useful in the management of hemobilia after blunt hepatic trauma and iatrogenic injury. These case reports and small series demonstrated that angioembolization of the liver could be used to successfully manage hemobilia. The operative management of hepatic trauma, whether the mechanism is blunt or penetrating, has evolved over last 15 years as well. Historically, the operative mortality for grade IV and grade V liver injuries has been reported as between 50% and 80%. About 10 years ago, the concept of damage control laparotomy was developed to try and avoid the “triad of death”: hypothermia, acidosis, and coagulopathy. This concept revolves around minimizing the length of the initial laparotomy in severely injured patients so that they arrive at the ICU for further resuscitation as soon as possible. This technique involves packing liver injuries rather than performing extensive operative maneuvers to gain hemostasis, repairing vascular injuries, and resecting bowel injuries, but leaving the bowel in discontinuity and temporary abdominal closure. This operative strategy has been reported to lower the operative mortality rate of grades IV and V liver injuries to 25%–40%. Angioembolization has been successfully used as an adjunct to damage control laparotomy to further decrease mortality and achieve hemostasis. Today, most authors would recommend a damage control laparotomy with packing of the injured liver for severe (AAST grade IV and V) hepatic trauma with angioembolization as an adjunct to further achieve hemostasis rather than extensive operative attempts to gain hemostasis. Angioembolization is also used as an adjunct to the nonoperative management of blunt hepatic injuries. Those patients who are seen to have active contrast extravasation from a liver injury on CT scan can undergo angioembolization to control their hemorrhage. This approach has improved the success rate of nonoperative management of blunt hepatic trauma to over 85%. The complications of hepatic embolization include hepatic necrosis, hepatic abscess, and bile leaks. If the patient has an intact portal vein with hepatopetal flow, the risk of hepatic necrosis is markedly reduced. For victims of both penetrating and blunt hepatic trauma, angioembolization can be a very useful adjunct to control the hemorrhage associated with significant liver trauma regardless of whether the patient is managed operatively or nonoperatively.

Angioembolization and endovascular stents have also been used in the management of renal trauma. Blunt renal injuries with evidence of active contrast extravasation on CT scan have been managed with arteriography and embolization of associated renal artery branches to achieve hemostasis. Endovascular stents have been used to manage both blunt and penetrating injuries to the renal arteries. The kidney is very sensitive to warm ischemia and that restoration of blood flow to the kidney needs to occur within the first 4–6 hours after injury. Due to time spent in transportation from the scene, resuscitation and evaluation in the trauma room and possibly CT scan, it is often 2–3 hours after injury that the diagnosis of a renal artery injury is made. As a result, the operative kidney salvage rate for exploration of the renal artery has been reported to be 5%–10%. This rate is so low that some authors do not recommend attempting to revascularize a unilateral blunt renal artery occlusion. There have been several small series and case reports of using endovascular stents in blunt renal artery injuries, which suggest that the kidney salvage rate for this approach may be as high as 25%. It is important to remember that all of the patients in these studies were hemodynamically normal and did not require operative exploration for associated injuries. Figure 3 demonstrates a blunt dissection of the left renal artery after a motor vehicle collision. Figure 4 is a CT scan after placement of 17 mm × 6 mm renal artery stent with restoration of perfusion to the left kidney. Further study of the use of endovascular stents in the management of both blunt and penetrating renal artery injuries is clearly warranted.

Figure 3 Dissection of the left renal artery and pseudoaneurysm of the splenic artery.

Figure 4 Computed tomography scan after placement of 17 mm × 6 mm renal stent with restoration of perfusion to the left kidney.

Angioembolization has also been used to successfully manage retroperitoneal hematomas with evidence of active extravasation and is the treatment of choice for these injuries due to the complexity of attemptng to control these injuries operatively.

PELVIC TRAUMA

Angioembolization has become part of the first-line treatment for pelvic hemorrhage associated with severe blunt pelvic fractures such as open-book or wind-swept pelvic fractures. An external binder such as the T-pod or an external fixator is placed to restore the conformation of the pelvis to allow for tamponade of pelvic venous bleeding. If the patient responds to resuscitation and does not have an indication for operative exploration, angioembolization is then used to control any arterial bleeding. Arterial hemorrhage is identified in approximately 10%–20% of these patients with significant pelvic trauma. Subselective transcatheter embolization can be performed with either gelfoam fragments, which dissolve over 1–3 weeks, or metallic coils, which are essentially permanent. Usually, attempts are made to occlude the bleeding artery as distally as possible. However, in the setting of massive pelvic hemorrhage, it is sometimes necessary to occlude one or both internal iliac arteries with gelfoam to achieve hemostasis. Vertical shear pelvic fractures are associated with injury to the superior gluteal artery that can be diagnosed and managed by angioembolization of the superior gluteal artery. In addition to angioembolization for control of arterial hemorrhage associated with severe pelvic trauma there have been a few case reports of the use of endovascular stents to treat iliac vein injuries associated with blunt pelvic trauma. There have been case reports of using endovascular stents to repair penetrating injuries to the iliac artery and vein with good initial results.

EXTREMITY TRAUMA

Arteriography of the extremity, whether intraoperative or in the interventional radiology suite, can be used to evaluate the injured extremity. In penetrating trauma patients who have “hard” signs of vascular injury, which includes expanding hematoma, pulse discrepancy, and pulsatile bleeding, usually surgical exploration is all that is performed. However, for patients with multiple penetrating wounds to the same extremity, it may be desirable to perform arteriography to more precisely locate the injury. In blunt trauma patients with posterior knee dislocations, floating knees (distal femur fracture with tibial plateau fracture), supracondylar knee fractures and tibial plateau fractures, there is an increased incidence of blunt popliteal injury. If there is a pulse discrepancy in an extremity with one of these injuries, arteriography is usually performed prior to or at the time of operative exploration to confirm the diagnosis of blunt popliteal injury. This assumes that obtaining the arteriogram will not significantly delay appropriate treatment of the vascular injury. Duplex scanning is increasing in specificity and sensitivity for these injuries but the gold standard remains arteriography. Development of the multidetector CT scanner and resultant ability to perform CTA may ultimately replace traditional catheter arteriography.

MANAGEMENT OF LATE COMPLICATIONS

Abscess formation can occur after many of the injuries associated with blunt trauma. Intra-abdominal abscesses can occur after blunt or penetrating abdominal trauma, especially when there has been an associated injury to the bowel. If technically possible, these abscesses are often best managed by the placement of a percutaneous drain. These are usually placed under ultrasonography or CT guidance. The infected material should be sent for Gram stain and culture to allow for appropriate antibiotic therapy. The catheter is left in place until drainage stops, and there is evidence of resolution of the abscess cavity on CT scan or other radiographic test. Percutaneous catheters are usually not able to drain hematomas completely because of the viscosity of the hematoma. Bilomas and bile leaks can complicate hepatic trauma. Bilomas can be treated by percutaneous drainage and bile leaks are managed by endoscopically placed biliary stents. Occasionally, a bile leak will need to be managed by percutaneous transhepatic cholangiography and catheter placement.