Key Words:
abdominal aorta , thoracic aorta , pseudoaneurysm , stent-graft , axillary artery , subclavian artery , carotid artery , vertebral artery , femoral artery , transection , endovascular aortic occlusion balloon
Introduction
Endovascular techniques have become essential for the successful management of many elective and emergent vascular surgical cases. Although diagnostic angiography has always played a central role in the evaluation and management of vascular trauma, interventions such as stent-grafting and coil embolization are being used increasingly as treatment modalities. The use of endovascular interventions increased in frequency during the last decade as improved devices and experience with catheter-based therapies continued to evolve. In a review of the National Trauma Data Bank, Reuben et al found an increase from 2% in 1994 to 8% in 2003 in the use of endovascular interventions for vascular trauma. Compared with traditional surgical repair, endovascular stent-grafting for the repair of traumatic arterial injury offers the advantage of decreased morbidity, because a remote access site may be used, avoiding surgical dissection and lengthy operating times. Endovascular management of vascular trauma seems particularly appealing in the setting of the multiply-injured patient and in cases of injuries to the “watershed” areas between the trunk and extremities where proximal vascular control can be quite difficult. This chapter reviews current literature with regard to the endovascular management of extracranial carotid injury, blunt thoracic and abdominal aortic injury, axillosubclavian injuries, and extremity vascular injury. The topic of use of aortic occlusion balloons in trauma will be discussed as well.
Endovascular Management of Carotid Arterial Injury
Injuries to the distal internal carotid, to the proximal common carotid, and to the vertebral arteries are amenable to endovascular adjuncts to arrest hemorrhage, exclude dissections or pseudoaneurysms, or assist with open repair. In patients with soft signs of a cervical vascular injury and with hemodynamic stability, there is ample time for further evaluation. Arteriography may identify an intimal flap, dissection, pseudoaneurysm, complete or partial transection, or thrombosis. Penetrating injuries that require immediate operative intervention are not typically amenable to a purely endovascular approach, although adjunctive endovascular techniques may be used to support the standard open repair of these injuries. Endovascular therapy with bare-metal or covered stents have been reserved primarily for evolving dissections that are surgically inaccessible, for pseudoaneurysms that persist after antithrombotic treatment, or for patients with worsening neurologic symptoms.
Neck injuries management has been classically divided according to three anatomic zones. Zone I injuries occur below the cricoid cartilage. Those presenting with hard signs of vascular injury may have an enlarging hematoma at the thoracic inlet, high chest-tube output, or hemodynamic collapse. These injuries notoriously involve the great vessels. Immediate control involves a high anterior thoracotomy, a sternotomy, or a clavicular resection to obtain adequate proximal control. An endovascular approach to injured neck arteries is appealing in that some of these injuries can be remotely approached from within the thoracic aorta thus avoiding the morbidity of extensive surgical exposure. Once the patient is prepared in the operative theater, an occlusion balloon can be used from access in the femoral vessels to provide endoluminal proximal control of the great vessels, which allows the conducting of a surgical exposure in a more controlled fashion and possibly avoiding sternotomy for proximal exposure. With an occlusion balloon in place, an arteriogram can locate the injury and can allow for operative planning. After the injury is exposed, a vascular clamp can replace the occlusion balloon; or, if proximal vessel length is not adequate, the occlusion balloon could remain in place during repair. In cases of blunt injury with a large mediastinal hematoma ( Fig. 19-1 ), the unique advantages for an endovascular approach include avoidance of a general anesthetic and the ability to monitor neurologic status during the intervention.
Zone II injuries have been classically managed with immediate exploration and direct evaluation of the aerodigestive tract and the carotid and jugular vessels. Direct examination and determination of hard signs predict those patients with significant injuries who might benefit from immediate exploration.
Zone III injuries above the angle of the mandible at the base of the skull are often difficult to expose and to control. Many case reports have demonstrated decreased morbidity and mortality for endovascular repair of traumatic injury to large vessels in this region using covered stents or endovascular coiling of posttraumatic pseudoaneurysms.
While there is not a comparative study of endovascular repair versus open repair, there are several case series in the literature reporting success with an endovascular approach. Initial reports of this application relied on balloon expandable Palmaz stents for pseudoaneurysms with and without dissections. Parodi and colleagues reported exclusion of 3 pseudoaneurysms with self-expandable covered stents, and 2 were treated with an autogenous vein-covered Palmaz stent. A vein graft theoretically reduces platelet aggregation and the potential for infection in grossly contaminated wounds. Covering the injury site promotes thrombosis of the pseudoaneurysm; but, if the sac fails to thrombose, one option is to coil-embolize the sac through the interstices of the bare stent.
Now that covered stents are commercially available in smaller diameters, a majority of pseudoaneurysms can be treated with self-expanding covered stents for both blunt and penetrating trauma. Du Toit et al examined a series of 19 zone I and zone III penetrating carotid injuries treated with stent-grafts over a 10 year period. The technical success rate was 100% with 1 patient experiencing a stroke within 30 days of the procedure. Of the 14 patients who had a mean follow-up of 44 months, there were no stent-graft–related strokes or deaths. Only one asymptomatic occlusion was detected on follow-up. DuBose et al reviewed 31 studies that examined stent placement for carotid artery injuries between 1994 and 2007. The postprocedural stroke rate for 113 patients was 3.5%, the occlusion rate was 10%, and the leak rate was 5% for a follow-up period of 2 weeks to 2 years. Cox et al treated 10 pseudoaneurysms from military injuries in the carotid and vertebral arteries with no neurological morbidity; however, 1 of the 2 stent-grafts occluded during follow-up. Cothren and colleagues reported their 3- to 6-month follow-up analysis of posttraumatic carotid pseudoaneurysms treated with carotid stents. Patients with a persistent pseudoaneurysm 7 to 10 days after injury were considered candidates for stent therapy. In the analysis, 23 patients were treated with Wallstents (Boston Scientific, Natick, MA), and 3 of those patients (13%) experienced ischemic complications: 2 periprocedural and 1 attributed to medication noncompliance. Edwards and associates placed 22 carotid stents for BCVI: 18 patients had pseudoaneurysms, and 4 patients were treated for extensive dissections. There were no periprocedural complications. Twelve patients in this series were treated with postprocedural antiplatelet therapy, and eight received anticoagulation. With a mean angiographic follow-up of 7 months, there were no occlusions (100% patency).
Follow-up is imperative in these patients as complications of concern include thrombosis, clinical embolic events, and stent fracture, though rarely reported. In a long-term follow-up of patients who underwent stenting for carotid artery dissection treated with uncovered stents, no patient experienced radiographic thrombosis;, and ischemic symptoms did not develop. Compliance with medications and follow-up surveillance should also be considered when planning appropriate therapy for trauma patients. Based on an extrapolation of data from carotid artery stenting for atherosclerotic disease, a regimen of dual antiplatelet therapy (aspirin and clopidogrel) appears adequate to prevent stent thrombosis and embolic ischemic events.
Liu and colleagues treated 7 patients with anticoagulation for 8 weeks followed by long-term aspirin after stenting for carotid dissection; and, at 14 months, no thrombosis or neurologic events occurred. If antiplatelet therapy is discontinued, stent thrombosis and resultant stroke are inherent risks. Duane and colleagues treated a patient with aspirin and clopidogrel after placement of a Wallstent; and, secondary to cost, the patient refused to comply with clopidogrel therapy. At follow-up, the stent was thrombosed but the patient did not experience neurologic sequelae. Anecdotally, most authors recommend some form of antiplatelet regimen for at least 6 weeks following stent implantation in a carotid artery based on carotid artery stenting for atherosclerotic disease. Additional prospective studies with long-term follow-up are needed to determine the risks and efficacy of carotid stents for BCI.
Endovascular Management of Vertebral Artery Injury
Vertebral artery injury is rare, but the identification of vertebral artery injuries has increased owing to the liberal use of screening tests and improved imaging during trauma workup. There are no data to support routine stenting for blunt vertebral arteries injuries; however, endovascular treatment of the vertebral artery with a combination of embolization techniques has been reported. This is usually in the setting of uncontrollable hemorrhage, arteriovenous fistulas, and pseudoaneurysm formation, as well as in cases of symptomatic patients who cannot tolerate anticoagulation.
Vertebral artery injuries are most commonly due to penetrating trauma ( Fig. 19-2 ). In a series of 101 patients with traumatic vertebral artery injury only 6 patients were the result of blunt trauma, while the remainder were secondary to gunshot wounds and stab injuries. The series showed that 50% required postoperative angiography and embolization for clinical arteriovenous fistulas and ongoing bleeding. In 50% of the cases undergoing angiography, the injured vertebral artery was thrombosed requiring no treatment, while the remainder required embolization using a combination of coils and detachable balloons. Several patients may require a combined approach involving both open ligation and endovascular embolization. In the cases of intact vessels, the injury can be crossed from an antegrade approach, allowing embolization of both outflow and inflow (endovascular trap embolization).
Endovascular embolization appears to benefit those patients with injuries or low-flow arteriovenous fistulas within the cervical portion of the vertebral artery. Vertebral artery injuries within 2 cm of the origin or within a short distance of the posterior inferior cerebellar artery are poor endovascular candidates. High-flow arteriovenous fistulas should also be avoided because of the risk of coil migration beyond the lesion.
Endovascular Management of Blunt Thoracic Aortic Injury
There has been a recent shift toward endovascular repair of blunt descending thoracic aortic injury (BTAI). Multiple series with short- and midterm follow-up indicate that thoracic endovascular aortic repair (TEVAR) is a viable alternative to open repair for traumatic aortic injuries, and several studies have demonstrated reduced mortality and paraplegia rates with endovascular repair of BTAI compared with open repair. Thus the most recent clinical practice guidelines published by the Society for Vascular Surgery suggests that endovascular repair of traumatic thoracic aortic injuries be performed preferentially over open surgical repair or nonoperative management ( Fig. 19-3 ). Available thoracic stent-grafts were originally designed for aneurysmal disease and in the early experience they were used in an off-label manner. The three endoprostheses approved by the FDA for the treatment of thoracic aortic aneurysms are the Gore Thoracic Aortic Graft, approved in March 2005 (Gore & Associates, Flagstaff, AZ); the Talent Thoracic Stent Graft, approved in June 2008 (Medtronic, Minneapolis, MN); and the Zenith TX-2, approved in May 2009 (Cook Inc, Bloomington, IN). The only device currently approved by the Food and Drug Administration to treat BTAI is the Conformable GORE TAG Thoracic Endoprosthesis (Gore & Associates, Flagstaff, AZ), which was approved in November 2011.
The pitfalls of TEVAR for BTAI to date have been caused by use of devices not designed to address the specific needs of the trauma population which differs from the population with aneurysms. The devices fell short in terms of diameter size, compliance, and size of the delivery system, as the trauma patients are younger with a mean age of 40 years and have relatively smaller aortic diameters, have smaller radius of aortic curvature or so called “Gothic” arches, and have smaller caliber access vessels.
An analysis of the angiographic morphology of 50 traumatic aortic disruptions showed that the mean aortic diameter adjacent to the region of injury was approximately 19.3 mm, smaller than the smallest available endoprosthesis, which measures 22 mm, but falling within the recommended 6% to 19% oversizing criteria. In addition, the mean distance from the left subclavian artery was 5.8 mm, thus in most cases the left subclavian artery was covered to achieve a 2-cm seal zone. Due to the size limitations particularly in aortas with a diameter less than 22 mm, off-label use of extension cuffs from abdominal aortic endografts (e.g., Gore Excluder Aortic Extender cuffs [Gore & Associates, Flagstaff, AZ]) and peripheral vascular stents (e.g., Zenith AAA iliac leg extensions [Cook, Inc., Bloomington, IN]) have been used to treat BTAI. Graft oversizing in the small aortas can lead to device compression as shown in Figure 19-4 . Graft collapse can also occur due to lack of apposition of the proximal graft along the inner curvature of the aortic arch seen in the cases of tight Gothic arches leads to the “bird-beaking” phenomenon as shown in Figure 19-5 . The recent release of the Conformable GORE TAG Thoracic Endoprosthesis (Gore & Associates, Flagstaff, AZ) device addresses the aortic size issue and allows treatment of patients with aortic diameters of 16 mm to 42 mm.
The large and bulky delivery devices ranging between 18 Fr and 25 Fr also pose an access challenge in a younger patient population with the small caliber iliofemoral vessels placing them at an increased risk for iatrogenic injury. Iliofemoral vessels that are smaller than 7 mm are associated with increased risk of iatrogenic complications, and creation of an iliac or aortic conduit may avoid potential iatrogenic injury.
Left arm ischemia has been reported following coverage of the left subclavian by the endograft when proximal landing zone extension is required. When this complication is encountered, it can usually be remedied with a carotid subclavian bypass. A recent study investigating the consequence of left subclavian artery coverage identified 94 studies incorporating 1704 patients with thoracic aneurysms and demonstrated that total left subclavian artery coverage without revascularization increases the prevalence of left arm ischemia (4% versus 0%); stroke (1.2% versus 0.23%); and the need for an additional procedure (2.86% versus 0.86%). In contrast, there were no reported cases of stroke, spinal cord ischemia, endoleak, stent migration or mortality when the left subclavian artery origin was only partially covered. There are ongoing investigations regarding the use of fenestrated endografts for preserving left subclavian artery during TEVAR for BTAI.
Paraplegia resulting from compromise to the collateral circulation of the spinal cord is reported in open repair of traumatic aortic disruptions but has been conspicuously absent in multiple metaanalyses where TEVAR for BTAI has been studied.
Repair Versus Observation
Repair is dictated by the type of injury. and the timing of repair depends on the patient’s associated injuries. Intimal tears (<10 mm) heal with nonoperative management. The University of Washington clinical treatment guidelines for blunt aortic injury are as follows:
- •
All patients with radiographic evidence of blunt aortic injury (BAI) should undergo antiimpulse therapy with b-blockade, if tolerated, coupled with antiplatelet therapy (81 mg aspirin).
- •
Observation alone with interval follow-up computed tomography angiography (CTA) within 30 days is appropriate for all intimal tears <10 mm.
- •
Selective management of large intimal flaps (>10 mm) is appropriate with repeat imaging within 7 days to assess for progression. Evidence of progression should be managed, when possible, with endovascular repair.
- •
All patients with an aortic external contour abnormality should be considered for semielective (<1 week) endovascular repair if there is a high likelihood of survival from other associated injuries. These patients should be monitored with CT imaging as follows: 1 month, 6 months, 1 year, and every other year thereafter. Patients with hypotension on presentation and with aortic arch hematoma >15 mm should be repaired with endovascular methods on a more urgent basis.
- •
Intentional left subclavian artery coverage without revascularization is well tolerated in a majority of patients with BAI.
- •
Patients with traumatic brain injury and an aortic external contour abnormality should be considered for earlier repair if a deliberate increase in mean arterial pressure is deemed beneficial for the patient.
Intravascular ultrasound use at the time of repair offers a useful adjunct to characterize the aorta proximal and distal to the injury site as well as to accurately measure the aortic diameter in the hyperdynamic aorta of the trauma patient ( Fig. 19-6 ).
