Surgical and endovascular repair of blunt thoracic aortic injury

INTRODUCTION — Blunt aortic injury is a potentially life-threatening condition that will require repair in patients with anything more than a minimal injury and is second only to traumatic brain injury as a cause of death in injured patients.

The most significant surgical advance in the past 30 years for the treatment of thoracic aortic disease has been endovascular stent-grafting [1]. Compared with other etiologies that are treated with thoracic endovascular stenting, such as thoracic aneurysm, patients with blunt aortic injury tend to be younger, with anatomic features that may present challenges for the placement of endograft devices, which were not originally designed for this indication, though later-generation devices are being designed to address these issues.

The surgical and endovascular options for repair of the blunt thoracic aortic injury will be reviewed. The clinical features, diagnosis, and management of blunt thoracic aortic injury are discussed separately. (See "Clinical features and diagnosis of blunt thoracic aortic injury" and "Management of blunt thoracic aortic injury".)

INDICATIONS FOR AND TIMING OF REPAIR — Aortic repair for blunt aortic injury is indicated for patients with blunt aortic injury grades II, III, and IV. The diagnosis, grading, and indications for repair, as well as timing of repair, are discussed elsewhere. Aortic repair is generally not indicated for patients with type I injury (intima only), as these can usually be successfully managed nonoperatively with optional interval imaging surveillance, but repair may become necessary if the injury progresses. When necessary, aortic repair in grade II or III patients may be delayed mitigating high surgical risk due to comorbidities, challenging aortic anatomy, or coexistent injuries. (See "Management of blunt thoracic aortic injury", section on 'Approach to management'.)

Issues to consider — There is preference for endovascular repair when anatomically suitable and when institutional resources are available to support the endeavor [2]. In a review of the National Trauma Data Bank (2003 to 2013), endovascular repair has nearly replaced open repair [3]. Nevertheless, for some patients, endovascular repair may not be a clear choice. An individualized approach accounts for the patient's specific injuries, age, anatomy, and other comorbidities. Open surgery, which is durable and does not require ongoing graft surveillance, may be a better option for some patients with an acceptable surgical risk and is the only option for patients with anatomic features that preclude endovascular repair. (See "Management of blunt thoracic aortic injury", section on 'Approach to management' and "Management of blunt thoracic aortic injury", section on 'Aortic repair'.)

Clinical and anatomic features that favor open surgical repair include:

●Anatomic features that are unsuitable for endovascular repair (see 'Anatomic criteria for endograft repair' below)

●Lesion involving the ascending aorta or aortic arch

●Need for open thoracic surgery to treat other injuries

Clinical and anatomic features that favor endovascular repair include:

●Multiple severe injuries

●Severe right chest or lung injuries; the patient may not tolerate one-lung ventilation, which is required for open surgery

●Multiple medical comorbidities

●Limited life expectancy

●Lesion confined to the descending thoracic aorta

ENDOVASCULAR REPAIR — Aortic endovascular repair was first described for the treatment of abdominal aortic aneurysm in 1991 and has been extended to the treatment of blunt aortic injury and other thoracic aortic pathologies [4-6]. The number of patients treated with endovascular repair of blunt traumatic injuries has increased dramatically. In the American Association for the Society on Trauma (AAST)-1 study (2008), stent-graft repair was selected in 65 percent of cases of aortic trauma [7,8]. A 2016 review of the National Inpatient Sample (NIS) in the United States showed that among patients who were managed operatively, 75 percent were treated using an endovascular stent-graft [9].

Endovascular repair for blunt thoracic aortic injury is performed in a similar manner to endovascular repair of thoracic aortic aneurysm, which is described in detail elsewhere. Technical success rates for blunt aortic injury vary, ranging from 80 to 100 percent [10]. (See "Endovascular repair of the thoracic aorta".)

The information provided below highlights differences in thoracic endograft placement for grafts placed in normal-caliber aortas for a variety of thoracic aortic pathologies, along with specific considerations in patients with thoracic aortic injury.

Anesthesia for endovascular repair is discussed separately. (See "Anesthesia for endovascular aortic repair".)

Endografts and sizing — Endovascular repair of blunt thoracic injury is accomplished using a fabric-covered stent termed an endograft or stent-graft. With improvements in stent-graft design, morbidity related to stent-graft placement for blunt aortic injury has decreased [11]. The essential features of these devices, and studies demonstrating their safety and efficacy, are discussed in detail elsewhere. (See "Endovascular devices for thoracic aortic repair".)

Thoracic endograft devices were originally designed for the treatment of nonruptured aneurysms of the thoracic aorta. The listed devices have technical advantages and disadvantages, and all are associated with technical success rates approaching 100 percent when used to treat focal lesions located within the descending portion of a nontortuous thoracic aorta [12,13]. However, thoracic stent-graft devices were not specifically designed for this indication, and 20 to 30 percent are not suited to the narrow, acutely angled aortic arch typical of the younger patient who presents with blunt aortic injury [8,9,14-16]. While early reports of long-term outcomes using these devices have been promising, these may be subject to bias due to issues with loss to follow-up. (See 'Long-term outcomes' below.)

Smaller-diameter thoracic aortas (≤17.5 mm) cannot be accommodated by any of the available thoracic endograft devices with the appropriate degree of oversizing to ensure proximal fixation. It is important to note that in hemodynamically unstable patients, the size of the aorta will likely increase once the patient is fully resuscitated [17]. In one study, the average aortic diameter in patients with blunt aortic injury was approximately 23 mm with a range of 19 to 30 mm [18]. A next-generation device, the Conformable TAG graft (CTAG), has been approved in the United States for the treatment of blunt aortic injury and was designed to conform to smaller, more tortuous, and/or tapered thoracic aortic anatomy to allow treatment of nonaneurysmal pathologies [19]. The CTAG is available in a minimal 21 mm diameter, which can accommodate an aorta that is 17.5 mm with 20 percent oversizing [20]. Some have suggested that oversizing the endograft for this indication is not as crucial as with aneurysmal disease, and that 10 percent oversizing should be adequate [21], particularly given that it appears that the aorta tends to dilate in response. (See 'Graft surveillance' below.)

An alternative approach uses abdominal aortic devices [18,22]. However, these devices are shorter in length, requiring more than one device for adequate coverage, and the delivery system is also shorter in length. None of the abdominal devices are suited to the tight angulation of the aortic arch in the younger patient.

Anatomic criteria for endograft repair — The following criteria are mandatory for thoracic endograft placement:

●Aortic diameter at proximal landing zone ≥17.5 mm

●Adequate proximal and distal landing zones (≥2 cm in length)

●Minimal calcification at the fixation sites

●Focal aortic lesion

●Location in the descending thoracic aorta

●No significant tortuosity of the thoracic aorta

●Minimal intraluminal thrombus

●Adequate diameter and quality of access vessels (sheath size depends upon the graft chosen)

Spinal drainage — The indications for specific measures to protect the spinal cord are similar to thoracic endovascular repair for aneurysmal disease. However, for blunt aortic injury occurring in the proximal portion of the descending thoracic aorta, spinal drainage is generally not needed. Spinal fluid drainage should be considered if the entire thoracic aorta requires endograft coverage, there has been a preexisting repair of the infrarenal aorta, or the area of coverage will include the distal third of the thoracic aorta. Alternatively, spinal drainage can be instituted if neurologic symptoms are detected during the postoperative period. (See "Endovascular repair of the thoracic aorta", section on 'Spinal cord ischemia'.)

Device placement — The patient is placed in a supine position. On rare occasions, conversion to an open thoracotomy may be needed, and thus preemptive placement on a bean bag (used for thoracotomy positioning) minimizes the delay. Arterial cutdown of the common femoral artery is usually performed, though a percutaneous approach is increasingly being performed [23,24]. Once the common femoral artery is exposed, the access sheath is placed into the common femoral artery and a guidewire is advanced through the sheath under fluoroscopic guidance into the ascending aorta. If the femoral artery is judged too small for the stent-graft delivery system (via imaging or on table assessment), iliac or aortic conduits can be used; however, conduits in the pelvis should be avoided in patients with pelvic fractures or pelvic hematomas [21].

Aortography can be performed via a catheter placed percutaneously through the contralateral groin (or brachial artery) to confirm the anatomy and define the landing zones for the device. Intravascular ultrasound (IVUS) is used by many endovascular surgeons to define and localize the injury and size the aorta and the device. Between the timing of the initial CT scan and device placement, the aortic diameter may have changed due to hemodynamic changes and resuscitation. IVUS provides real-time information to confirm device sizing at the time of placement to avoid potential undersizing based on a CT performed when the patient may still have been in shock. Transesophageal echocardiography (TEE) can also be used to provide similar information to aid guidewire advancement and graft deployment [25].

The device is positioned across the area of pathology by advancing it over a stiff guidewire (eg, Lunderquist). The position of the device is confirmed with repeat aortography. Prior to device deployment, the patient's blood pressure should be lowered to lessen the arterial impulse and minimize the potential for misplacement during deployment. The device is then deployed, thus excluding the site of injury from the circulation. A completion aortogram should be performed to confirm patency of the endograft and successful exclusion of the area of disruption (ie, no endoleaks). IVUS can also be used to evaluate the repair after device deployment. TEE can also provide accurate identification of early endoleaks (sensitivity 0.875, 95% CI 0.51-0.99; specificity 0.67, 95% CI 0.28-0.90). In a single-center study, contrast TEE was more accurate than conventional TEE and angiography for the detection of endoleaks, incomplete stent expansion, further need for balloon inflation, or further stent implantation [26]. Once the positioning and integrity of the repair is confirmed, the guidewires and introducer sheaths are removed, and the femoral arteriotomies are repaired.

Anticoagulation — Systemic anticoagulation with heparin is typically used during the placement of endovascular grafts to reduce the risk of thrombus formation within the thoracic aorta and around the sheath site, which could lead to distal embolization and possibly stroke. For the patient with a contraindication to heparin, an alternative agent can be used. (See "Management of heparin-induced thrombocytopenia (HIT) during cardiac or vascular surgery" and "Management of heparin-induced thrombocytopenia", section on 'Choice and dose of non-heparin anticoagulant'.)

However, for patients with traumatic aortic injury, significant associated injuries, such as intracranial hemorrhage or splenic injury, may contraindicate anticoagulation. Under these circumstances, endovascular repair of the thoracic aorta without the use of heparin has been used without adverse consequences [27,28]. For patients with traumatic aortic injuries and a contraindication to anticoagulation (eg, major intracranial injury), particularly if the aortic injury appears to be in an anatomically favorable location, we proceed with endograft repair without systemic heparinization. This approach may also allow earlier intervention.

Subclavian coverage — An adequate proximal landing zone may require coverage of the left subclavian artery, particularly in patients with blunt aortic injury, in whom the injury is typically located at the level of the aortic isthmus. However, positioning the top of the endograft proximal to the origin of the left subclavian artery can lead to poor apposition of the graft along the inner curvature of the aortic arch, particularly in younger patients who have a narrower aorta. This may increase the risk of endoleak or endograft collapse. (See 'Endograft-related complications' below.)

In our experience, complete or partial coverage of the origin of the left subclavian artery is needed in approximately 20 percent of cases, but the need can be higher among those with complicated injuries [18,29]. Although coverage of the subclavian artery origin is usually well tolerated, symptoms of vertebrobasilar insufficiency can develop in patients with a dominant left vertebral artery, and coronary steal syndrome can develop in patients who have a left internal mammary coronary bypass graft. These patients may benefit from extra-anatomic bypass (carotid subclavian bypass) prior to coverage of the left subclavian with the stent-graft [18]. (See "Endovascular repair of the thoracic aorta", section on 'Arch vessel bypass' and "Surgical and endovascular techniques for aortic arch branch and upper extremity revascularization", section on 'Arterial bypass'.)

A systematic review identified 94 studies that evaluated subclavian coverage during endovascular repair of blunt aortic injury in 1704 patients [30]. There were no cases of left upper extremity ischemia, stroke, spinal cord ischemia, endoleak, or death when the left subclavian artery was revascularized, or when the left subclavian origin was only partially covered. By contrast, total coverage of the left subclavian artery without revascularization increased the incidence of left arm ischemia (4 versus 0 percent), stroke (1.2 versus 0.2 percent), and the need for an additional procedure (2.9 versus 0.9 percent). In a prospective study, patients presenting with blunt aortic injury and a <2 cm distance between the site of rupture and the left subclavian artery underwent intentional coverage [31]. Among 31 patients, 9 developed upper extremity ischemic symptoms, which was significantly related to left vertebral artery diameter <3 mm. The majority of patients with ischemic symptoms had significant symptomatic improvement; overall, two patients required reintervention at 48 months of follow-up.

An alternative to subclavian coverage is the use of fenestrated grafts, originally described for repair of thoracic aortic aneurysm. Several studies report good outcomes for patients with blunt aortic injury [32,33]. (See "Endovascular repair of the thoracic aorta", section on 'Need for debranching procedures'.)

Postoperative care and follow-up — Following endovascular repair, the patient should be monitored in the intensive care unit setting. The access site should be checked for signs of bleeding, and distal pulses serially examined. Once the endograft is in place, the heart rate and blood pressure can be maintained in a normal range. Neurologic assessment should be performed at routine intervals in the first 24 hours to evaluate for neurologic complications of thoracic aortic repair such as paraplegia or stroke.

If the patient has no other associated injuries or medical conditions requiring treatment, they can be discharged typically within a couple of days of the procedure, and normal activities can typically be resumed within a couple of weeks.

Following successful aortic repair, beta blockers or other antihypertensive therapy instituted upon admission can be weaned off. There is no need for these if not previously indicated for other reasons (eg, essential hypertension).

Graft surveillance — Following endograft placement, the aorta dilates at the site of implantation [34,35]. Several studies have investigated the alterations in aortic diameter that occur as a result of endograft placement for blunt aortic injury. One retrospective study compared pre-and postprocedure imaging in 70 injured patients who received a CTAG device (see 'Endografts and sizing' above) with 54 patients who underwent endovascular aneurysm repair (EVAR) for aneurysm [34]. The proximal and distal aortic neck demonstrated remodeling with a positive correlation between the increase in aortic diameter and the amount of endograft oversizing. Mean increases in proximal and distal neck diameters between pretreatment and 30-day imaging were significantly greater in traumatic injury patients than in aneurysm patients. In both study populations, smaller pretreatment aortic neck diameters showed a larger change in neck diameter compared with larger pretreatment aortic diameters, which some speculate is due to undersizing related to hypotension. These studies support the need for ongoing endograft surveillance.

The optimal program for thoracic endograft surveillance is still being defined. We suggest obtaining repeat imaging (usually contrast-enhanced computed tomography [CT] of the chest) prior to hospital discharge, within a few days of the repair. Thereafter, routine surveillance is typically performed in the postoperative period (eg, one month, six months), then annually, but routine surveillance varies by institution. Thereafter, routine surveillance is typically performed in the postoperative period (eg, one month, six months), then annually, but routine surveillance varies by institution [21]. The need for ongoing follow-up is important when selecting patients for endovascular repair, particularly in the trauma population, and, for some patients, an inability to comply with graft surveillance may be a deciding factor favoring open surgical repair. Loss to follow-up remains an issue following both open and endovascular repair in this population of patients. (See 'Issues to consider' above and 'Long-term outcomes' below.)

The need for lifelong graft surveillance in patients who have undergone endovascular repair for blunt aortic injury has been challenged by some [36,37]. Over the course of time, radiation exposure may become a significant issue, particularly in young patients who will be subjected to numerous CT scans over their lifetime. For patients who have demonstrated stable-appearing endograft repairs, the interval between studies can be lengthened to reduce the risks associated with repeated radiation exposure. Alternatives to chest CT are magnetic resonance (MR) angiography [36,37] and transesophageal echocardiography [38]. (See "Radiation-related risks of imaging".)

OPEN SURGICAL REPAIR — Patients with blunt thoracic aortic injury are typically younger with no significant preexisting medical comorbidities; however, coexisting injuries, particularly traumatic brain injury and lung injury, as well as other medical issues may influence the choice and timing of repair. (See "Clinical features and diagnosis of blunt thoracic aortic injury" and "Management of blunt thoracic aortic injury", section on 'Immediate versus delayed'.)

Open repair of traumatic aortic injury is similar to open repair of thoracic aortic aneurysm with some notable exceptions, which are discussed below. Open repair of thoracic aortic aneurysm is discussed in detail elsewhere. (See "Management of thoracic aortic aneurysm in adults".)

Open surgical repair of the thoracic aorta is performed under general anesthesia using a double-lumen endotracheal tube. One-lung ventilation allows collapse of the left lung and facilitates access to the aorta. It is important to note that patients with severe associated right chest or lung injuries may not tolerate one-lung ventilation. Delayed repair or endovascular repair may be warranted in these patients. (See "Overview of open surgical repair of the thoracic aorta" and "One lung ventilation: General principles".)

The extent of the thoracotomy incision is based upon the location of the pathology. As examples, in blunt aortic injury, the aortic disruption usually occurs at the aortic isthmus, and thus, a fourth intercostal space thoracotomy incision should be performed. More extensive involvement may require sternotomy or thoracoabdominal exposure.

Once the chest is entered, the site of pathology is identified and any bleeding should be controlled with direct pressure while an aortic clamp is placed just distal to the left subclavian artery. If the aortic disruption is in close proximity to the left subclavian artery, the clamp will need to be placed between the left common carotid artery and the left subclavian artery. In some cases, such as a small traumatic tear, the aorta can be repaired primarily; however, most patients require the placement of an interposition graft (eg, Dacron).

Perfusion techniques — Clamping the thoracic aorta compromises the blood supply to the spinal cord. Active perfusion of the distal aorta, which is a source of flow to the spinal arteries, has reduced the incidence of paraplegia associated with repair of blunt thoracic aortic injury to approximately 2 percent, with some centers reporting no spinal cord complications [39-44]. (See 'Spinal cord ischemia' below.)

In the early experience of open thoracic aortic repair for blunt thoracic aortic injury, most procedures were performed with a clamp-and-sew technique. However, even a short cross clamp time (<30 minutes) did not prevent spinal cord ischemia and neurologic deficits [8,45-47]. A meta-analysis reported a 19 percent incidence of paraplegia and mortality of 16 percent [39]. An older technique, passive perfusion, in which a heparin-bonded shunt is placed to divert blood from the proximal aorta above the site of clamping to the distal aorta below the site of the clamping, has been supplanted by active perfusion techniques [39].

Active perfusion requires anticoagulation, typically using heparin, but an alternative agent may be needed. (See "Blood management and anticoagulation for cardiopulmonary bypass", section on 'Systemic anticoagulation'.)

Active perfusion can be achieved using one of two techniques:

●Left atrial-femoral bypass – Left atrial-femoral bypass is accomplished by placing a cannula into the left atrium (or pulmonary vein). The oxygenated blood is pumped and delivered to the distal aorta via a cannula placed into the left common femoral artery. This simple circuit can be performed with a relatively small dose of heparin (approximately 5000 units). One of the disadvantages of this technique is that it relies on oxygenation of the blood in a patient using a single lung. The level of oxygenation may not be sufficient in patients with concomitant pulmonary contusion.

●Femoral venous-to-femoral arterial cardiopulmonary bypass – An alternative technique places a cannula in the femoral vein and advanced to the right atrium. A second cannula is placed in the femoral artery. This technique is known as femoral venous-to-femoral arterial cardiopulmonary bypass (venoarterial bypass). With this technique, deoxygenated blood is drawn from the cannula in the right atrium and circulated through a cardiopulmonary bypass machine to oxygenate the blood before returning it to the patient via the femoral artery. Advantages of the venoarterial bypass include:

•It allows simultaneous cooling of the patient, which may augment spinal cord protection.

•Oxygenation of the blood is independent of the lung, which is preferred in patients with concomitant lung injury [43].

•Systemic heparin administration can be minimized by using heparin-coated tubing [42].

In one 30-year retrospective review, paraplegia did not occur in 73 patients who underwent repair of the thoracic aorta for traumatic injury using venoarterial bypass [48]. In another retrospective review, 138 consecutive patients were treated over a 35-year period and repaired with a short interposition graft using some form of circulatory support [49]. The average age was 27 years, and 91 percent had associated injuries (mean injury severity score [ISS] 44). Death occurred in 7/138 patients (5 percent). Mortality varied with method of circulatory support and was 7.5 percent in patients perfused with the Gott shunt and 4 percent (4/98) in those with left heart bypass. Among the most recent 68 cases using left heart bypass, the mortality rate was lower (1.5 percent).

PERIOPERATIVE MORBIDITY AND MORTALITY — Perioperative (30 day) outcomes after endovascular repair appear to be improved compared with open surgery [10,50-54]. However, the long-term effect of endovascular stenting in this predominantly younger population is yet unknown.

The American Association for the Society on Trauma (AAST)-1 study evaluated 274 cases of blunt aortic injury from 50 trauma centers over a two-and-a-half-year period [8]. A variety of open surgical techniques were used, and overall mortality was 31 percent. Approximately 10 years later, the AAST-2 study found a modest reduction in mortality to 23.5 percent for patients undergoing open surgical repair [7]. During that time period, thoracic endovascular grafting was introduced, and mortality associated with it was significantly lower at 7.2 percent compared with open repair [55]. In the RESCUE study (endovascular aortic repair using Valiant Captivia for blunt thoracic aortic injury), a nonrandomized multicenter study, perioperative (30 day all-cause) mortality was 8 percent (4 of 50 patients), and one-year all-cause mortality was 12 percent (6 of 50 patients) [56,57].

A systematic review compared the outcomes of 7768 patients drawn from 27 retrospective studies and 112 case series who were treated with endovascular or open repair and, in some cases, with no repair [19]. Overall mortality was lower for patients who underwent endovascular compared with open repair or no repair (open repair: 9 versus 19 percent; no repair: 9 versus 46 percent). Injury severity score correlated with mortality for open surgical repair, but not for endovascular repair.

Complications — The main complications associated with open and endovascular repair of the thoracic aorta for blunt aortic injury include access complications, stroke, spinal cord ischemia, and endograft-related complications. Although less common, late complications related to open repair can occur (eg, coarctation, aneurysmal degeneration).

Spinal cord ischemia — Spinal cord ischemia may result in paraparesis or paraplegia. Although the incidence of spinal cord complications has steadily diminished with the increased use of active perfusion techniques, paraplegia remains a significant concern when open repair is used to repair the thoracic aorta. The incidence of paraplegia ranges between 0 and 9 percent [7,19,55]. The AAST-2 study reported an overall reduction in the incidence of paraplegia from 8.7 to 1.6 percent (open and endovascular repair). This reduction was correlated to a 20 percent increase in the use of spinal cord perfusion techniques [55]. The incidence of paraplegia associated with endovascular repair decreased to 0.8 percent over that same time period. In a systematic review that included 7768 patients undergoing open surgical or endovascular repair for thoracic aortic injury, the risk of spinal cord ischemia was significantly lower for endovascular (3 versus 9 percent) versus open repair [19].

Stroke — Perioperative stroke is a potentially devastating complication of repair of blunt aortic injury. The risk of stroke is higher for repair of aortic injury that is proximal to the origin of the left subclavian artery.

In a review of five investigational endovascular device trials that included 60 patients with blunt aortic injury, the incidence of cerebrovascular events (excluding transient ischemic attack) was 10 percent [58]. An earlier systematic review of 33 retrospective studies included 370 patients who underwent endovascular repair and 329 open repairs for blunt aortic injury [10]. Stroke occurred in 1 percent of patients following endovascular repair and 4.5 percent of patients undergoing open surgical repair [10]. The details of the injuries treated were not provided in the review. Other reviews of thoracic endovascular repair performed for assorted thoracic aortic pathologies report stroke rates ranging from 0 to 4.5 percent [59-62].

In observational studies, coverage of the left subclavian artery during endograft placement increases the reported incidence of stroke during thoracic endovascular aneurysm repair [63-65]. In one of these studies, the incidence of stroke increased with subclavian coverage both with and without left subclavian revascularization compared with no left subclavian coverage (with revascularization: 4.7 versus 2.7 percent; without revascularization: 4.1 versus 2.6 percent) [64]. A similar comparison has not been performed for blunt aortic injury.

Endograft-related complications — Endovascular repair of the thoracic aorta is associated with a 20 percent incidence of graft-related complications including technical failure, endoleak, stent-graft collapse, iliac artery injury, and other device-related failures (eg, metal stent penetration, device component disconnect) [18,66-70]. Improvements in the design of thoracic endovascular grafts that take into account the anatomy of the nonaneurysmal aorta, particularly in younger patients, should help reduce the incidence of device-related complications. In one retrospective review, freedom from intervention at five years was estimated at 96 percent [71].

Most endografts that have complications can be salvaged with endovascular techniques such that conversion to open repair is not necessary.

Device complications can be related to endograft undersizing, excessive oversizing, or related to device deployment. As an example, positioning the endograft with its leading edge perpendicular to blood flow in the transverse aortic arch can lead to device collapse [70,72-74]. Efforts to avoid oversizing can lead to inappropriate undersizing. Undersizing may also be the result of an underestimation of the diameter of the aorta in the face of systemic hypotension. The highly elastic thoracic aorta recoils, leading to measurements on computed tomography (CT) that can underestimate the true aortic diameter at normal blood pressure [17]. Intravascular ultrasound during endovascular repair is useful for confirming aortic diameter in real-time as the patient is being resuscitated [8,75-77].

Endoleak in the nonaneurysmal aorta is defined as persistent flow around the graft in the native aorta following endovascular repair. In the AAST-2 study of blunt aortic injury, the incidence of endoleak (mostly type I) was 14.4 percent, of which one-third subsequently required open repair [7]. Our experience has been similar. In 43 consecutive cases, six endoleaks occurred with three patients requiring open surgical repair and endograft explantation [18]. (See "Endovascular repair of the thoracic aorta", section on 'Evaluating for endoleak'.)

LONG-TERM OUTCOMES — In spite of the short-term benefits of decreased mortality and paraplegia, the long-term durability of thoracic endograft devices for the repair of nonaneurysmal pathologies is not certain. Many patients with blunt aortic injury are young, and it is unknown how the endograft will respond to normal aging. The thoracic aorta tends to enlarge with age, but it is uncertain just how the aorta adjacent to the endograft responds and whether any subsequent aortic dilation will result in endoleak. (See 'Graft surveillance' above.)

In a review of 60 patients (mean age 43 years) with a mean follow-up of five years, there were no returns to the operating room for aortic-related problems or stent-graft failure [78]. Overall survival was 87 percent at 1 year, 82 percent at 5 years, and 75 percent at 10 years. In a smaller review with follow-up to 12 years, none of 11 patients had any major complications [79]. In a multi-institution series of 74 patients, overall five-year survival was 82 percent, with one-half of deaths occurring perioperatively or in the first year following injury [80].

In a review of Taiwan's National Health Insurance Database, 287 patients were surgically managed (open, endovascular) for thoracic aortic injury over a 10 year period [81]. The proportion of survivors was 71.9 percent in the open surgery group and 88.9 percent in the endovascular group at one year; 68.2 and 88.9 percent at three years, respectively; and 65.1 and 88.9 percent at five years, respectively. Up to five years follow-up, the proportion of freedom from reintervention was 100 percent in the open surgery group and 97.4 percent in the TEVAR group. The endovascular group had two events of late reintervention and one event of late cerebrovascular accident.

Although long-term outcomes appear promising, loss to follow-up is a problem with this population of patients [82-84]. Loss to follow-up ranges from 0 to 57 percent at one year for endovascular repair, and 0 to 28 percent for open repair, worsening over longer follow-up intervals [84]. As an example, in one study, the percentages of patients who were followed at one, three, and five years from operation were only 62, 25, and 14 percent, respectively [82]. In another, stent-graft surveillance computed tomography scans were not obtained in 37.6 percent [83].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Thoracic trauma".)

SUMMARY AND RECOMMENDATIONS

●For patients in whom repair is indicated for blunt thoracic aortic injury, and who have anatomy that is suitable for thoracic stent-graft placement, endovascular rather than open surgical repair is recommended. Endovascular repair is associated with significantly decreased mortality and a lower incidence of spinal cord ischemia compared with open surgical repair. However, the long-term outcomes of thoracic endovascular repair, particularly in younger patients, are not known. Reintervention following endovascular repair is needed in up to 20 percent of patients. Thus, patients who undergo endovascular repair require lifetime surveillance, which results in the associated risks of repeated radiation exposure. (See 'Indications for and timing of repair' above and 'Perioperative morbidity and mortality' above.)

●Endovascular repair of blunt thoracic injury is accomplished using a fabric-covered stent termed an endograft or stent-graft. Narrow thoracic aortas (≤17.5 mm) cannot be accommodated by any of the available, dedicated thoracic endograft devices with the appropriate degree of oversizing needed to ensure proximal fixation. Efforts are ongoing to optimize thoracic endovascular grafts for the treatment of blunt thoracic aortic injury, including provision of smaller-diameter stent-grafts and more flexibility to accommodate the more acutely angulated aortic arch of younger individuals. (See 'Endografts and sizing' above and "Endovascular devices for thoracic aortic repair".)

●Although endovascular repair may be preferred, patients with blunt thoracic aortic injury have multiple issues that must be considered when determining the appropriate timing and method of repair. These include the presence and type of associated injuries, anatomic features such as the location of the injury and diameter of the aorta, the patient's life expectancy, and the likelihood of compliance with long-term follow-up with endovascular repair. (See 'Issues to consider' above.)

●The major complications associated with both open and endovascular repair of the thoracic aorta for blunt aortic injury are spinal cord ischemia and stroke. The risk of spinal cord ischemia is lower for endovascular compared with open repair of blunt thoracic aortic injury. Comparative stroke rates are less well defined. For endovascular repair, endograft complications occur in approximately 20 percent of patients and include access-related vascular injury, technical failure in placement, and other device-related problems such as endoleak, stent-graft collapse, stent-graft migration, and device component disconnect, among others. (See 'Complications' above.)