Endovascular repair of traumatic aortic transection

ARTICLE IN PRESS doi:10.1510/icvts.2010.235473

Interactive CardioVascular and Thoracic Surgery 11 (2010) 238–242 www.icvts.org

Institutional report - Aortic and aneurysmal

Endovascular repair of traumatic aortic transection Eleftherios Chalvatzoulisa, *, Angelos Megalopoulosa , George Trellopoulosa , Olga Ananiadoua, Pavlos Papoulidisa, Ioanna Kemanetzib, Athanasios Madesisa, George Drossosa Department of Cardiothoracic and Vascular Surgery, ‘George Papanikolaou’ General Hospital, Exohi, 57010 Thessaloniki, Greece b Department of Cardiac Anesthesia, ‘George Papanikolaou’ General Hospital, Thessaloniki, Greece

a

Received 16 February 2010; received in revised form 27 May 2010; accepted 28 May 2010

Abstract The present study reports a single center experience and mid-term results of endovascular repair of acute aortic traumatic transection on an emergency basis, instead of open surgical management. From January 2005 to December 2008, 13 cases of traumatic aortic transection with serious comorbidities, which underwent repair with thoracic stent grafts at our institution, were reviewed. The mean patient age was 32.5"7.8 years. During the follow-up period of 25.5"12.8 months, 12 patients were alive and one patient died of associated injuries. There were no intraoperative deaths, no incidence of paraplegia and no procedure-related mortality. The blood loses during the procedure were minimal. The thoracic aortic grafts were larger than the thoracic aorta by 12.4"5.7%. Five cases required complete or partial coverage of the left subclavian artery. There were two cases of graft collapse, which were successfully treated by endovascular reintervention. Our results suggest that this approach is safe, effective and can be performed with low rates of morbidity and mortality, especially in respect of patients with multiple injuries. Although initial results are encouraging, close long-term follow-up and technical improvements of the stent grafts are required. 䊚 2010 Published by European Association for Cardio-Thoracic Surgery. All rights reserved. Keywords: Thoracic aortic trauma; Endovascular treatment

1. Introduction Traumatic aortic transection (TAT) is a potentially lethal injury with out of hospital mortality of 86.2%. It is second only to head injury as the most common cause of death following blunt trauma, with an incidence between 13% and 20% w1x. TAT results from rapid deceleration and application of shearing forces and the site of aortic tear is usually at the isthmus of the aorta w1, 2x. Despite recent advances in surgical and anesthetic techniques, surgery for TAT is still associated with significant morbidity and mortality. Postoperative paraplegia is the main neurological complication, with a rate of occurrence ranging from 2.3% to 25.5% w3x. The use of stent grafts for the treatment of abdominal aortic aneurysmal disease has encouraged many investigators to apply the same principles in the treatment of TAT w4x. Although endovascular stent graft technologies, have been increasingly employed as an emergency treatment for TAT with good improved mid-term mortality and morbidity compared to conventional open surgery, the reported series remain small and the long-term consequences of this treatment remain unknown w5x. A number of controversies have been generated concerning their use, such as early vs. delayed treatment, coverage of the left subclavian artery *Corresponding author. Tel.: q30 2313 307661; fax: q30 2313 307667. E-mail address: [email protected] (E. Chalvatzoulis). 䊚 2010 Published by European Association for Cardio-Thoracic Surgery

(LSA)yleft common carotid artery, degree of safe oversizing in small aorta patients, device flexibility in acute aortic arch, use as first line treatment in very young patients, and device related complications (endoleak, stent collapse, migration). In this paper, we report our experience with endovascular repair of blunt injury to the thoracic aorta in 13 patients, we focus on technical aspects and evaluate the problems and failure modes of these devices. 2. Material and method 2.1. Patient selection From January 2005 to December 2008, 13 patients underwent endovascular repair of TAT. The mean patient age was 32.5"7.8 years (range, 18–45 years) and all were male. Nine patients suffered motorcycle trauma, four suffered a fall from a height and the cause of injury in all cases was blunt trauma from rapid deceleration. In all cases, an endovascular approach was the treatment of choice because the presence of serious comorbidities and associated injuries made open thoracic aortic surgery extremely invasive and high-risk (Table 1). The diagnosis of TAT was established with the use of computed tomographic (CT) angiography, after clinical sus-

ARTICLE IN PRESS E. Chalvatzoulis et al. / Interactive CardioVascular and Thoracic Surgery 11 (2010) 238–242

239 Editorial

Table 1. Patient characteristics Interval from diagnosis to repair (h)

Associated injuries

1

M

23

Motor vehicle accident

-24

2 3 4 5 6

M M M M M

18 25 27 30 40

Motor vehicle accident Motor vehicle accident Motor vehicle accident Fall from height Fall from height

-12 -24 -24 -24 -24

7 8 9 10 11 12 13

M M M M M M M

37 32 45 36 37 31 41

Motor vehicle accident Fall from height Motor vehicle accident Motor vehicle accident Fall from height Motor vehicle accident Motor vehicle accident

-12 -12 -12 -12 -24 -12 -12

Splenic rupture, intraabdominal hemorrhage, upper and lower limb fractures Pulmonary contusions, multiple rib fractures Pulmonary contusions, multiple rib fractures, upper limb fractures Pulmonary contusions, multiple rib fractures, lower limb fractures Fractured pelvis, lower limb fractures Liver laceration, intraabdominal hemorrhage, fractured pelvis, lower limb fractures Liver laceration, intraabdominal hemorrhage Fractured pelvis, lower limb fractures Lower limb fractures Lower limb fractures, hemothorax Pulmonary contusions, subdural hematoma Pulmonary contusions, multiple rid fractures, subdural hematoma Pulmonary contusions, subdural hematoma, fractured pelvis, internal iliac artery rupture

M, Male; y, year; h, hour.

Table 2. Diagnostic findings and aortic measurements Distance from LSA (mm)

Proximal aortic neck diameter (mm)

Distal aortic neck diameter (mm)

1 2 3 4 5 6 7 8 9 10 11 12 13

Complete laceration False aneurysm, periaortic hematoma False aneurysm False aneurysm False aneurysm Complete laceration False aneurysm False aneurysm Complete laceration, periaortic hematoma Complete laceration, hemothorax False aneurysm False aneurysm False aneurysm

– 40 38 42 41 – 32 52 – – 46 50 37

23 15 28 23 41 19 24 14 40 28 20 27 21

21 21 24 26 28 24 26 26 34 26 20 22 23

18 19 22 24 27 23 25 25 33 25 19 21 21

LSA, left subclavian artery.

Brief Case Report Communication

False aneurysm diameter (mm)

Historical Pages

Diagnostic findings

Nomenclature

Patient no.

Best Evidence Topic

A plain chest radiograph was obtained on the first to third postoperative day as a record of the stent graft skeleton

State-of-the-art

Stent graft sizing was based on CT angiography and angiographic images. Procedures were performed in the

Follow-up Paper

2.3. Follow-up

Negative Results

2.2. Endovascular technique

Proposal for Bailout Procedure

angiography suite and operating theatre and cardiopulmonary bypass facilities were available for conversion to open repair if required. Five procedures were performed under regional anesthesia and eight patients received general anesthesia, based on clinical need. All patients were systematically heparinized with 5000 IU heparin, except for cases with head injuries. The standard operative approach involved an open cut down of one common femoral artery with a left brachial artery sheath inserted and advanced into the LSA and aortic arch for proximal forward-flow aortography. DSA was performed to reveal the anatomy of the aortic arch. The vascular access was problematic in three patients forcing us to use the device without its sheath. The delivery system was inserted under fluoroscopic control and the stent graft was released to the appropriate position, confirmed by DSA through the LSA catheter. On completion, DSA confirmed good graft position and detected the presence of endoleaks.

ESCVS Article

picion was raised by symptomatology, chest radiograph findings, or the mechanism of injury. CT angiography also provided preoperative evaluation of cerebral circulation and in two less demonstrable cases supplemental intraarterial digital subtraction angiography (IA DSA) revealed cerebrovascular anatomy and also confirmed the diagnosis of TAT. The CT findings showed complete aortic laceration in four patients and false aneurysm localized in the area of the isthmus in nine patients. Distance between the lesion and the LSA ostium was 24.8"8.2 mm (range, 14– 41 mm) and the proximal aortic neck diameter was 24.7"3.7 mm (range, 20–34 mm) (Table 2). Seven patients were treated within 12 h and six within 24 h of diagnosis. Six patients were operated prior to the endovascular procedure, due to intra-abdominal hemorrhage, multiple fractures or subdural hematoma and nine patients had orthopedic or vascular surgery after the stent placement (Table 1).

Institutional Report

Mechanism of injury

Protocol

Age (y)

Work in Progress Report

Sex

New Ideas

Patient no.


240

and position. Follow-up imaging involving CT angiography, was routinely performed at first, sixth and 12th month after the procedure and annually thereafter. Nevertheless, we conducted more frequent CT-scan evaluation in a few patients based on symptomatology, i.e. thoracic pain postprocedurally. 3. Results The mean operative time was 46.5"9.2 min. The blood loss during the procedure was 178"122 ml (Table 3). There was no conversion to open thoracotomy and we had no incidence of paraplegia postoperatively. In our series, there was no intraoperative death and no stent graft-related deaths. One patient (no. 12) died the second postoperative day in the intensive care unit (ICU), due to extensive cerebral hemorrhage. In cases with systematic heparinization, there was no incidence of augmented bleeding from associated injury sites. Also, using the minimum amount of contrast media needed, no renal function deterioration was observed. The mean aortic diameter was 24.7"3.7 mm, whereas the stent graft diameter was 27.6"3.2 mm (range, 24– 36 mm) and the length was 107.7"18.8 mm (range, 100– 150 mm). The grafts were oversized by 12.4"5.7% (range, 5.9–23.8%) compared with the normal, adjacent proximal aorta. Fourteen devices were used in total, seven TALENT (Medronic Vascular, Santa Rosa, CA, USA) and seven GORE TAG (W.L. Gore & Associates, Flagstaff, AZ, USA) (Table 3). In all 13 cases, completion angiography demonstrated secure exclusion of the traumatic isthmic aortic lesion. In patient 8 a second stent was deployed due to a detected endoleak, resulting in exclusion of the false aneurysm and complete occlusion of LSA ostium due to the short distance with the injury level. In all consecutive cases, transection was detected in the descending thoracic aorta within 41 mm of the LSA and in nine patients there was at least 20 mm of normal aorta between the LSA and transection. Five patients experienced cessation of antegrade flow through the LSA without developing left upper extremity symptoms or suffering from cerebrovascular accident. Particularly, patients 7, 11 and 13 suffered LSA partial coverage and patients 2 and 8

complete coverage. In these two patients the systolic pressure gradient between arms was 70 mmHg at the early postoperative phase. They remained asymptomatic and the pressure gradient was found to be 20 mmHg, in the oneyear follow-up. Furthermore, partial LSA coverage was well tolerated with no development of type IA endoleaks or vessel thrombosis. We recorded two cases of stent collapse. In patient 1, the aorta proximal to the injury was measured as 21 mm with 23 mm being the distance from the LSA. A 26 mm TAG endoprosthesis (oversizing 23.8%) was used to repair the transection (Table 3). Although the GORE TAG device is designed to be oversized from 7% to 18% (http:yywww. goremedical.comyenyfileyAK0314.pdf), we were forced to use a 26 mm graft due to devices of smaller size being unavailable at our institution at that time. Stent grafting was successful, without occlusion of LSA ostium. Four days later he developed acute renal failure, pulmonary edema and unpalpable pulse of both femoral arteries. CT demonstrated a proximal graft collapse, which was successfully treated with reintervention and new in-stent placement (TAG 26 mm) that covered the LSA ostium, leaving the patient without any symptoms or signs of ischemia of the upper limp, although initially a systolic pressure gradient of 80 mmHg was observed between the arms. Renal and respiratory function improved gradually and the patient was discharged from ICU on the eight postoperative day. Patient 6 suffered complete laceration at the level of the isthmus (Fig. 1) and underwent a stent graft (TAG 26 mm) placement with 8.33% oversizing. He underwent CT angiography four days later due to acute thoracic pain and a decrease in the hematocrit level, which revealed incomplete attachment of the stent graft to the lesser curvature of the aortic arch with a portion of the proximal stent protruding in the aortic arch (Fig. 2). The patient was pain free after receiving systematic analgesia, he remained asymptomatic during hospitalization and no further action was taken. The follow-up CT scan in the 45th postoperative day revealed stent graft collapse (Fig. 3). That was treated with reintervention and an additional thoracic endoprosthesis (TALENT 28 mm) completely covered the LSA ostium (Fig. 4).

Table 3. Stent graft and endovascular procedure details Patient no.

Device type

Device dimensions

Graft oversizing (%)

Number of grafts used

LSA ostium

Blood loses (ml)

1 2 3 4 5 6 7 8 9 10 11 12 13

GORE TAG TALENT GORE TAG GORE TAG GORE TAG GORE TAG TALENT TALENT GORE TAG GORE TAG TALENT TALENT TALENT

26 24 28 28 31 26 28 28 36 28 24 26 26

23.80 14.28 16.66 7.69 10.71 8.33 7.69 7.69 5.88 7.69 20 18.18 13.04

1 1 1 1 1 1 1 2 1 1 1 1 1

Patent Covered Patent Patent Patent Patent Partly covered Covered Patent Patent Partly covered Patent Partly covered

500 350 70 100 150 100 100 200 200 100 100 200 150

LSA, left subclavian artery.

mm=100 mm=100 mm=100 mm=100 mm=100 mm=100 mm=100 mm=100 mm=100 mm=100 mm=100 mm=150 mm=150

mm mm mm mm mm mm mm mm mm mm mm mm mm


241 Editorial New Ideas Work in Progress Report Protocol Institutional Report

Fig. 1. Computed tomography angiography showing complete laceration of the descending thoracic aorta distal to the origin of the left subclavian artery.

Negative Results

During the follow-up period (25.5"12.8 months, range 6– 47 months), 12 patients were alive, with no evidence of leak, or distal migration.

Follow-up Paper State-of-the-art

Despite significant advances in critical care medicine and refinement in surgical techniques morbidity and mortality rates of open surgical repair of TAT are still high. Postoperative paraplegia is of major concern with an incidence up to 25.5% and 30-day mortality ranging from 8 to 30% w3x. Some authors have advocated that open repair remains the first line therapy for blunt aortic injury in patients -18 years of age and those with an outer wall aortic diameter -18 mm for reasons of oversizing but in many centers it has now become the initial procedure of choice, even in young or low-risk patients. As for the timing of repair, in the case of aorta-related hemodynamic instability (massive mediastinal hematoma, active bleeding, or left hemothorax), emergency endovascular treatment must be performed. In the case of non-aorta-related hemodynamic

Proposal for Bailout Procedure

4. Discussion

ESCVS Article

Fig. 3. Routine follow-up computed tomography angiography at 45th postoperative day revealing stent graft collapse.

Best Evidence Topic Nomenclature Historical Pages

Fig. 4. Computed tomography angiography showing new instent placement that covered completely the ostium of the left subclavian artery.

Brief Case Report Communication

Fig. 2. Computed tomography angiography four days postoperatively demonstrating the incomplete attachment of the proximal part of the stent graft to the lesser curvature of the aortic arch.

ARTICLE IN PRESS 242

E. Chalvatzoulis et al. / Interactive CardioVascular and Thoracic Surgery 11 (2010) 238–242

instability, the other life-threatening injuries should be treated first, and endovascular treatment of the aortic injury can be performed within 24 h w6x. Endovascular repair of TAT has clear advantages over open surgery, avoiding left thoracotomy, single lung ventilation, systemic heparinization, and cardiopulmonary bypass, in often critically injured patients. Low paraplegia incidence is also due to the relative short length of endografts w5x. Furthermore, as seen in our study, operative times and mean blood loss are low and technical success is good with no endovascular graft related mortality. Intentional covering of LSA may lead to higher incidence of extremity ischemia and stroke. The evidence in the literature is sparse with a low number of events and the outcomes are poorly described in many cases although it needs to be weighed against many factors, such as urgency of repair or patient anatomy. In our series, it was well tolerated without prophylactic surgical transposition or the need for intervention in the follow-up period w5, 7x. Another potential technical problem is the vascular access in young patients in shock because a minimum diameter of 6 mm is required for safe access. We report the successful use of a TAG device without its sheath in three cases, taking into account the risk of stent deployment in the incorrect position. Rates of stent graft collapse in a series of endovascular transection repairs range from 0.03% to 10% w5, 7, 8x. Anatomical factors, such as the acute angle of the aortic arch, together with excessive oversizing, can lead to graft collapse w9, 10x. Commercially available stent grafts have large diameters, mandating excessive oversizing in patients with TAT, with smaller sized aorta, although collapse without excessive oversizing has also been reported w11x. For this reason, some authors are proposing the use of small diameter abdominal aortic endograft cuffs or iliac limbs w12x. The more appropriate diameter of these cuffs, however, is counter-balanced by the fact that they are shorter devices with short device shaft lengths. Furthermore, the tighter curvature of the aortic arch in young trauma patients often results in an incomplete graft apposition at the inner curve. As a result, the graft is exposed to the pulsatile blood flow, with twice the peak blood flow velocities than in older patients w13x. In this regard, Yamaguchi and colleagues report their early experience with no collapse or stent graft fracture with placement of curved flexible devices in treating TAT w14x. As for the durability requirements for stent grafting in the acute setting, these may include a three-fold longer time period than for older aneurysm patients. In our limited follow-up period we did not record any case of deteriorating prosthetic materials or mechanical failure of the stent graft.

Consistent with our results, in a recent retrospective review, Urgnani and colleagues report promising short- and mid-term results of endovascular repair of acute traumatic aortic injuries w5x. Furthermore long-term survival studies are mandatory to determine the efficacy and durability of this technique. In addition, the engineering and technical aspects of thoracic stent grafts designed for use in blunt aortic injury are likely to be improved in the next few years. Delivery devices of smaller profile, more flexible devices, precurved reinforced grafts and also fenestrated stent grafts that preserve blood flow through branches of the aortic arch, would be beneficial to apply this technique in thoracic aortic trauma. References w1x Smith RS, Chang FC. Traumatic rupture of the aorta: still a lethal injury. Am J Surg 1986;152:660–663. w2x Sevitt S. The mechanism of traumatic rupture of the thoracic aorta. Br J Surg 1977;64:166–173. w3x Attar S, Cardarelli MG, Downing SW, Rodriguez A, Wallace DC, West RS, McLaughlin JS. Traumatic aortic rupture: recent outcome with regard to neurologic deficit. Ann Thorac Surg 1999;67:957–964. w4x Thompson CS, Rodriguez JA, Ramaiah VG, DiMugno L, Shafique S, Olsen D, Diethrich EB. Acute traumatic rupture of the thoracic aorta treated with endovascular stent graft. J Trauma 2002;52:1173–1177. w5x Urgnani F, Lerut P, Da Rocha M, Adriani D, Leon F, Riambau V. Endovascular treatment of acute traumatic thoracic aortic injuries: a retrospective analysis of 20 cases. J Thorac Cardiovasc Surg 2009;138:1129– 1138. w6x Marcheix B, Dambrin C, Bolduc JP, Arnaud C, Hollington L, Cron C. Endovascular repair of traumatic rupture of the aortic isthmus: midterm results. J Thorac Cardiovasc Surg 2006;132:1037–1041. w7x Rizvi AZ, Murad MH, Fairman RM, Erwin PJ, Montori VM. The effect of left subclavian artery coverage on morbidity and mortality in patients undergoing endovascular thoracic aortic interventions: a systematic review and meta-analysis. J Vasc Surg 2009;50:1159–1169. w8x Go MR, Barbato JE, Dillavou ED, Gupta N, Rhee RY, Makaroun MS, Cho JS. Thoracic endovascular aortic repair for traumatic aortic transection. J Vasc Surg 2007;46:928–933. w9x Canaud L, Alric P, Branchereau P, Marty-Ane ´ C, Berthet JP. Lessons learned from midterm follow-up of endovascular repair for traumatic rupture of the aortic isthmus. J Vasc Surg 2008;47:733–738. w10x Muhs BE, Balm R, White GH, Verhagen HJ. Anatomic factors associated with acute endograft collapse after GORE TAG treatment of thoracic aortic dissection or traumatic rupture. J Vasc Surg 2007;45:655–661. w11x Steinbauer MG, Stehr A, Pfister K, Herold T, Zorger N, To ¨pel I, Paetzel C, Kasprzak PM. Endovascular repair of proximal endograft collapse after treatment of thoracic aortic disease. J Vasc Surg 2006;43:609– 612. w12x Peterson BG, Matsumura JS, Morasch MD, West MA, Eskandari MK. Percutaneous endovascular repair of blunt thoracic aortic transection. J Trauma 2005;59:1062–1065. w13x Salmasi AM, Dore ´ C. Variation of aortic blood velocity with age at rest and during exercise in normal subjects. Clin Auton Res 1995;5:19–23. w14x Yamaguchi M, Sugimoto K, Tsukube T, Mori T, Kawahira T, Hayashi T, Nakamura M, Kawasaki R, Sandhu R, Sugimura K, Kozawa S, Okita Y. Curved nitinol stent-graft placement for treating blunt thoracic aortic injury: an early experience. Ann Thorac Surg 2008;86:780–786.