Endovascular Repair of Blunt Traumatic Thoracic Aortic Injuries: Seven-Year Single-Center Experience

Hypothesis  Thoracic endovascular aortic repair (TEVAR) for acute blunt thoracic aortic injury has good early and mid-term results.

Design  Single-center retrospective 7-year review from January 2001 to December 2008.

Setting  Urban tertiary care hospital.

Patients  Twenty-four consecutive patients with acute blunt thoracic aortic injury treated with TEVAR.

Main Outcome Measures  Procedure-related mortality, stroke, or paraplegia; injury severity score; and complications.

Results  Among the 24 treated patients (mean age, 41 years; range, 20-71 years), the mean injury severity score was 43 (range, 25-66). Thoracic endovascular aortic repair was successful in treating the aortic injury in all patients and there were no instances of procedure-related death, stroke, or paraplegia. Access to the aorta was obtained through an open femoral/iliac approach (n = 7) or an entirely percutaneous groin approach (n = 17). Systemic heparin was not used in 84% of cases. Two access complications (8%) occurred, requiring an iliofemoral bypass in one patient and a thrombectomy in another. One patient required secondary intervention for device collapse, which was treated successfully with repeat endografting. There have been no delayed device failures or complications among the entire cohort at mid-term follow-up.

Conclusion  Thoracic endovascular aortic repair, via a percutaneous groin approach and without systemic anticoagulation, for blunt thoracic aortic injury can be performed safely with low periprocedural mortality and morbidity.

Blunt thoracic aortic injury is a potentially devastating event with a reported 85% prehospital mortality.¹ This injury represents the second most common cause of death due to blunt trauma.² A recent review of the National Trauma Data Bank found an overall blunt thoracic aortic injury incidence of 0.3% in 1.1 million trauma admissions in a 5-year period. Sixty-eight percent of the patients who survived transport and triage underwent no repair, with an associated mortality rate of 65%.³ Typically, blunt thoracic aortic injury is associated with significant concomitant injuries, including closed head injury, pulmonary contusion, long-bone and pelvic fractures, and solid organ injury, increasing the overall morbidity and mortality and making the use of systemic anticoagulation problematic.⁴

During the last 50 years, open repair of blunt thoracic aortic injury has been the standard of care, with mortality rates ranging from 5% to 28% and paraplegia rates secondary to spinal cord ischemia ranging from 2.3% to 14%.^2,4 Definitive treatment is often delayed owing to the need to manage multiple associated injuries; this approach carries a risk of an in-hospital rupture rate of 2% to 13%.^4,5 The use of endovascular stent grafts as an expedient method to treat acute blunt thoracic aortic injury has the theoretical advantage of avoiding a thoracotomy, aortic cross-clamping, need for single-lung ventilation, and in some cases the use of systemic heparinization. The purpose of this study is to analyze our early and mid-term outcomes of thoracic endovascular aortic repair (TEVAR) to treat acute blunt thoracic aortic injury during the last 7 years at a single institution.

This study is a retrospective review of 24 consecutive patients who underwent TEVAR for blunt thoracic aortic injury from January 2001 to December 2008 at Northwestern Memorial Hospital in Chicago, Illinois. The facility, located in an urban setting, is a level 1 trauma center and performs a high volume of vascular and endovascular cases.

Abnormalities seen on initial chest roentgenograms in the setting of blunt trauma (including widened mediastinum, obliteration of the aortic knob, tracheal deviation to the right, left hemothorax, depression of the left main stem bronchus, presence of a pleural or apical cap, widened paratracheal stripe, or fractures of the first rib or scapula) initiated additional imaging workup with an intravenous contrast chest computed tomographic (CT) scan. This allowed for measurements of the proximal and distal neck length and aortic diameter as well as location and extent of injury in relation to the origins of the arch vessels.

During the time frame of this study, there were no cases of blunt thoracic aortic injury treated with an open repair in our institution. One patient underwent no repair secondary to nonsurvivable injuries.

This study was approved by the human rights committee and by the institutional review board at Northwestern University. Data management was performed solely by the Division of Vascular Surgery, whose members had full access to the data and vouch for the accuracy and completeness of the data and the analyses. Data analysis includes demographics, procedural records, radiographic imaging, and outcomes during a mean follow-up period of 21 months.

All procedures were performed in an operating room angiosuite by board-certified vascular surgeons with advanced endovascular skills. The devices used to treat the transections consisted of commercially available endovascular abdominal aortic proximal extension cuffs (AneuRx, Medtronic/AVE, Santa Rosa, California; and Excluder, W.L. Gore & Associates, Flagstaff, Arizona) in 18 patients and a Gore thoracic aortic graft (W. L. Gore & Associates) in the other 6 patients (Table 1).

After the induction of general anesthesia, access to the aorta was percutaneously obtained via the common femoral artery (17 patients) or by an open femoral/iliac artery approach (7 patients). A floppy-tipped J-wire was then passed into the aorta under direct fluoroscopic guidance. An aortogram was obtained using steep left anterior oblique projection. Arch anatomy and location of the aortic transection were noted. In patients undergoing percutaneous repair, a 10F Perclose device (Abbott Vascular, Redwood City, California) was then used with the previously described preclose technique.⁶ At this point, a delivery sheath was exchanged and passed proximally over a stiff wire. The endovascular stent graft device was then placed over the area of transection and deployed without the use of chemically induced cardiac arrest or hypotension. Additional aortic extension cuffs were placed as required to exclude the aortic injury.

After TEVAR, patients underwent CT angiography at 1, 6, and 12 months after repair and yearly thereafter. After 2 years of treatment, CT imaging was reduced to every 2 to 3 years. Mean follow-up of this patient cohort is 21 months (range, 3-49 months).

Twenty-four patients underwent TEVAR for blunt thoracic aortic injury: 17 men (71%) and 7 women with a mean age of 41 years (range, 20-71 years). The mechanism of injury was motor vehicle crash in 19 patients, fall from a great height in 2, diving injury in 1, pedestrian vs automobile accident in another, and bicycle vs automobile in the last. Most of the patients had multiple injuries from blunt trauma and had a mean injury severity score of 43 (range, 25-66).

In eighteen patients (75%), the repair was performed within the first 24 hours of injury. Based on preoperative CT measurements, the mean native aortic diameter at the planned proximal landing zone was 23 mm (range, 19-30 mm).

Thoracic endovascular aortic repair was technically successful in 100% of patients, as determined by completion angiography at the time of intervention and CT scan (Figure). There were no instances of procedure-related death, stroke, or paraplegia. The overall 30-day mortality was 4% (n = 1). This patient died of multisystem organ failure as a consequence of other associated traumatic injuries. The mean device diameter was 28.5 mm (range, 23-34 mm). When abdominal aortic extension cuffs were used, 2 overlapping devices were used in 6 cases, 3 devices in 8 cases, and more than 4 devices in 4 cases.

Only 4 patients (16%) received intraoperative systemic heparin for the procedure. This subgroup of patients was treated more than 48 hours after the injury.

To ensure an adequate seal proximal to the transection, the origin of the left subclavian artery was completely covered with the stent graft in 7 patients, for a left subclavian artery coverage rate of 29%. In this group, the left subclavian artery was transposed to the left common carotid artery before endovascular repair in 6 patients, and in the other after 1 year secondary to upper extremity ischemic symptoms.⁷

Two vascular access site complications (8%) occurred, requiring an iliofemoral bypass in one patient and a thrombectomy in another. One patient required secondary intervention 1 day after treatment for device collapse that was treated successfully with repeat endografting.

The mean follow-up for survivors was 21 months. During follow-up, 1 patient died of unrelated causes 2 years after the initial trauma. At follow-up there have been no delayed device failures or complications (eg, endoleaks, migration, or late pseudoaneurysm formation) among the entire cohort.

Since the article by Semba et al,⁸ TEVAR has emerged as an alternative minimally invasive treatment option for blunt thoracic aortic injury. In the last decade there have been a number of case series demonstrating the acute feasibility of endovascular repair for blunt thoracic aortic injury (Table 2). The purported advantages of this treatment modality include the avoidance of thoracotomy, single-lung ventilation, aortic cross-clamping, cardiopulmonary bypass, and systemic heparinization. In 1997 the American Association of Trauma published its first prospective multicenter study for the treatment of blunt thoracic aortic injury (American Association for the Surgery of Trauma Study 1 [AAST1]). There were no endovascular interventions, and the authors reported an overall mortality of 31% and a postoperative paraplegia rate of 8.7%. The rate of death in patients who did not undergo surgery (not including patients arriving in extremis) was 55%.¹⁶ In 2008, the results of AAST2, a second prospective multicenter study, were reported. The study showed a trend toward selection of TEVAR (0% in AAST1 vs 64.8% in AAST2), a reduction in overall mortality, and a decrease in the incidence of procedure-related paraplegia from 8.7% to 1.6%. Although they are encouraging results, the authors also noted a 20% device-related complication rate and a 14.4% endoleak rate.^17,18 In a recent meta-analysis of stent graft repair for blunt thoracic aortic injury, Tang et al¹⁹ demonstrated a perioperative mortality of 7.6%, stroke rate of 0.8%, and paraplegia rate of 0%. Similarly, Xenos et al²⁰ reported in his analysis a procedure-related mortality of 2%, 30-day mortality of 8%, and paraplegia rate of 0%.

Most TEVAR procedures for blunt thoracic aortic injury are performed via a transfemoral approach, as was the case in our current series. However, we were successful in completing the TEVAR as an entirely percutaneous procedure using the previously reported preclose technique in most of our cases (71%).⁶ The reported benefits of this approach are fewer wound complications, decreased incidence of cutaneous nerve injuries, and a more expeditious treatment.^6,21,22

One of the primary concerns of open blunt thoracic aortic injury repair among patients with multiple injuries, particularly closed-head injuries, is the hemorrhagic complications associated with systemic intraoperative anticoagulation. While anticoagulation may still be needed when performing TEVAR for blunt thoracic aortic injury, lower doses are usually needed. Moreover, some centers have reported on successful outcomes of TEVAR for blunt thoracic aortic injury without the use of systemic anticoagulation.¹¹ In our current series, 20 patients (84%) did not receive systemic heparinization; of these, only 1 patient had a thrombotic complication that required an iliofemoral thrombectomy. Thoracic endovascular aortic repair without anticoagulation allows for more expedient repair of the acute aortic injury among patients with associated closed-head or long-bone injuries.

To date, no specific device has been approved for the treatment of blunt thoracic aortic injury. Nevertheless, off-label use of aortic endografts has incorporated a part of the armamentarium for treatment of blunt thoracic aortic injury. We first reported on the off-label use of proximal abdominal aortic extension cuffs in the treatment of traumatic aortic rupture several years ago.²³ In this current series, we used these devices in 18 patients. The advantage of this approach is the more appropriate diameter of these cuffs relative to the true aortic diameter. However, this advantage is counterbalanced by the fact that delivery systems for these cuffs lack sufficient length because the aortic cuffs were originally designed for infrarenal use.

As expected, endoleaks after TEVAR for blunt thoracic aortic injury is rather uncommon. We have demonstrated an endoleak rate of 0%. In general, trauma patients have a healthy, normal-caliber aorta with an acute tear, without degenerative aortic pathology. As such, the proximal seal zone of 2 cm, recommended to successfully treat thoracic aortic aneurysms, is probably not necessary when treating patients with blunt thoracic aortic injury. What has become a newly recognized concern of TEVAR for blunt thoracic aortic injury is the occurrence of device collapse.^24-26 It has been demonstrated that most blunt thoracic aortic injuries are observed in a younger patient population, which possess a more tightly angulated aortic arch. This factor, combined with excessive device oversizing relative to the normal aortic diameter, may predispose cases to device infolding and failure. We observed this complication in 1 patient, who was subsequently treated with relining of the Gore TAG endograft with abdominal aortic proximal extension cuffs.

The management of the left subclavian artery is a technical issue that remains a topic of interest. Although some patients tolerate this well, coverage of the left subclavian artery can result in posterior strokes or left upper extremity ischemia. In our institution, we favor revascularization of the left subclavian artery either with a subclavian carotid artery transposition or a bypass when preoperative imaging shows there is a possibility of complete left subclavian artery coverage to adequately exclude the injury and a dominant vertebral artery arises from the left subclavian artery.²⁷

Studies evaluating long-term outcomes are needed. The patient population with traumatic aortic ruptures is relatively young, and their life expectancy is considerable and exceeds the current experience with endovascular grafts.

This series demonstrates that the adaptation of commercially available stent graft devices to treat blunt thoracic aortic injury is technically feasible and can be performed entirely percutaneously with low rates of morbidity and mortality. A therapeutic approach without systemic anticoagulation allows a more timely treatment in the multiply injured patient without significant associated complications. The long-term durability of endovascular repair of blunt thoracic aortic injury remains unknown, but early and mid-term results appear promising.

Correspondence: Mark K. Eskandari, MD, Division of Vascular Surgery, Northwestern University, 676 N St Clair, Ste 650, Chicago, IL 60611 (meskanda@nmh.org).

Accepted for Publication: January 25, 2010.

Author Contributions:Study concept and design: Garcia-Toca, Naughton, Matsumura, Morasch, Kibbe, and Eskandari. Acquisition of data: Garcia-Toca, Naughton, Matsumura, Morasch, Pearce, and Eskandari. Analysis and interpretation of data: Garcia-Toca, Rodriguez, and Eskandari. Drafting of the manuscript: Garcia-Toca, Morasch, and Eskandari. Critical revision of the manuscript for important intellectual content: Naughton, Matsumura, Morasch, Kibbe, Rodriguez, Pearce, and Eskandari. Statistical analysis: Eskandari. Administrative, technical, and material support: Matsumura and Kibbe. Study supervision: Naughton, Morasch, Rodriguez, and Eskandari.

Financial Disclosure: None reported.

Previous Presentation: This paper was presented at the 117th Scientific Session of the Western Surgical Association; November 9, 2009; San Antonio, Texas; and is published after peer review and revision. The discussions that follow this article are based on the originally submitted manuscript and not the revised manuscript.

Benjamin Starnes, MD, Seattle, Washington: These authors report their modern experience with the endovascular management of blunt aortic injury in 24 patients over a 7-year period. Although a small series, this retrospective review emphasizes several important issues revolving around the management of these devastating injuries. Namely that endovascular repair has become the mainstay of therapy for blunt aortic injury with low associated rates of mortality, stroke, and paraplegia when compared with traditional open repair. There are no current FDA [Food and Drug Administration]-approved devices for the management of these injuries and off-label use of current devices is what is required for successful treatment. And that the perioperative use of heparin sulfate is not necessarily needed when these procedures are done carefully and rapidly using a percutaneous approach.

I have the following 3 questions for the authors:

1. Refinements in CT angiography have been spectacular over the past decade, so much so that injuries once missed with conventional imaging are being identified at a much higher rate. At Harborview in Seattle, we see an enormous volume of these injuries, some of which we believe don't need treatment, such as small intimal flaps without an associated pseudoaneurysm, that completely heal on follow-up imaging. What are your anatomic criteria for treating these injuries or do you treat everyone with an aortic injury? When should we intervene?

2. The authors only report their experience with TEVAR and fail to mention if any patients were treated with open repair at their institution over this 7-year period. Were any patients treated with open repair and why? What were the selection criteria for treating these patients and, more importantly, what were the outcomes?

3. Finally, it is widely known that the surgeons at Northwestern tout left subclavian artery revascularization for the management of thoracic aortic pathology when the left subclavian artery requires coverage during endovascular repair to mitigate the risk of stroke and paraplegia. At our own institution and with a 10-year series of over 60 patients treated with TEVAR for blunt aortic injury, we have found that the left subclavian artery can nearly always be covered with impunity and without the need for revascularization specifically for the management of these types of injuries. With regard to blunt aortic injury exclusively, what evidence do you have to convince us that this should be done routinely or is this approach purely speculative?

Dr Eskandari: Regarding when to intervene for these transections, not all injuries are of the same magnitude and our philosophy has been to treat all injuries with associated periadventitial hematomas, pseudoaneurysms, or dissections with intraluminal thrombus. As you point out, small intimal injuries can frequently be managed with blood pressure control alone. All of the cases presented today had significant injuries.

The second question is regarding open repair during this time period. Surprisingly, we have not had any open repairs performed since we started endovascular treatment for these injuries. Over the years, there has only been 1 patient we intended to treat endovascularly, but he died on his way to the operating room.

The last question is in regards to management of the left subclavian artery. There are several studies that suggest that complete coverage of the left subclavian artery during treatment for thoracic degenerative aneurysms is associated with an increased risk for posterior cerebral strokes and paraplegia. In the setting of acute trauma, I would agree with you that coverage of the left subclavian artery probably can be done in most cases with impunity. Certainly the caveats would be those individuals who on imaging have a dominant left vertebral artery or patients with a preexisting left internal mammary coronary artery bypass graft. Fortunately, most blunt thoracic aortic injuries can be successfully managed with thoracic stent graft that does not completely cover the left subclavian artery.

Financial Disclosure: None reported.