Endovascular management of traumatic ruptures of the thoracic aorta: A retrospective multicenter analysis of 28 cases in The Netherlands

Transcription

1 Endovascular management of traumatic ruptures of the thoracic aorta: A retrospective multicenter analysis of 28 cases in The Netherlands Liselot L. Hoornweg, MD, a Maarten K. Dinkelman, MD, b J. Carel Goslings, MD, PhD, c Jim A. Reekers, MD, PhD, d Hence J. M. Verhagen, MD, PhD, b Eric L. Verhoeven, MD, PhD, e Geert Willem H. Schurink, MD, PhD, f and Ron Balm, MD, PhD, a Amsterdam, Utrecht, Groningen, and Maastricht, The Netherlands Background: Minimally invasive endovascular treatment of a traumatic rupture of the thoracic aorta is a new strategy in the care of multitrauma patients. We report the experience in The Netherlands with endovascular management of patients with acute traumatic ruptures of the thoracic aorta. Methods: We reviewed 28 patients with a traumatic thoracic aortic rupture treated with a thoracic aortic endograft between June 2000 and April All patients underwent treatment at one of the four participating level 1 trauma centers. Data collected included age, sex, injury severity score, type of endovascular graft, endovascular operation time, length of stay, length of stay in the intensive care unit, and mortality. Follow-up data consisted of computed tomographic angiography and plain chest radiographs at regular intervals. Results: All patients (mean age, 40.9 years; SD, 18.5 years) experienced severe traumatic injury, and the mean injury severity score was 37.1 (SD, 7.8). All endovascular procedures were technically successful, and the median operating time for the endovascular procedure was 58 minutes (interquartile range, minutes). The overall hospital mortality was 14.3% (n 4), and all deaths were unrelated to the aortic rupture or stent placement. There was no intervention-related mortality during a median follow-up of 26.5 months (interquartile range, months). Postoperative data showed no severe endovascular graft- or procedure-related morbidity, except for one patient with an asymptomatic collapse of the endovascular graft during regular follow-up. This was corrected by placing a second graft. Conclusions: This study shows that the results of immediate endovascular repair of a traumatic aortic rupture are at least equal to those of conventional open surgical repair. Especially in these multitrauma patients with traumatic ruptures of the thoracic aorta, endovascular therapy seems to be preferable to conventional open surgical repair. (J Vasc Surg 2006; 43: ) Traumatic aortic rupture is a frequent cause of death in trauma patients. A total of 10% to 15% of severely injured victims of motor vehicle accidents die because of a traumatic aortic rupture. 1 Blunt trauma, especially high-speed motor vehicle accidents with a side or a frontal impact, and falls from great heights are well-known causes of a traumatic aortic rupture. The acute deceleration induces shear forces that tear off the relatively mobile aortic arch from the more fixed descending aorta at the level of the aortic isthmus. 2,3 As a result of this mechanism, most (90%) traumatic aortic ruptures are located at the isthmus, although ruptures may be found in various other parts of the thoracic aorta. 1 From the Departments of Vascular Surgery, a Trauma Surgery, c and Radiology, d Academic Medical Center, Amsterdam, the Department of Vascular Surgery, b University Medical Center, Utrecht, the Department of Vascular Surgery, e University Medical Center Groningen, and the Department of Vascular Surgery, f University Hospital Maastricht. Competition of interest: none. Reprint requests: Ron Balm, MD, PhD, Department of Vascular Surgery G4-107, Academic Medical Center, Meibergdreef 9, 1105 AZ Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands ( r.balm@amc.nl) /$32.00 Copyright 2006 by The Society for Vascular Surgery. doi: /j.jvs Traumatic rupture of the thoracic aorta is often fatal: only 9% to 20% of the patients arrive at the hospital alive. 1,3,4 Because these survivors have experienced a major accident, they often have multiple injuries, including head and neck injuries, blunt abdominal injuries, pulmonary contusions, and fractures of the limbs, spine, and pelvis. Although mortality rates of open repair of traumatic rupture of the thoracic aorta have improved over the last 5-8 decades, reported rates still vary between 8% and 19%. Because of this, it has been advocated to put patients on antihypertensive therapy and to delay surgery until patients are in a more stable pulmonary condition. However, even with careful blood pressure monitoring, some of the patients experience a total aortic rupture while awaiting stabilization and repair. 4,9,10 11 In 1991, Volodos et al described the first abdominal and thoracic endovascular grafting. The minimally invasive character of the procedure makes this technique especially suitable in high-risk patients with severe comorbidities and in multitrauma patients. Since the first reports on endovascular treatment of traumatic aortic ruptures in 1997, several small series have demonstrated endovascular repair to be an effective new treatment strategy with promising results The aim of this larger study was to report the experience in The Netherlands with acute endovascular

2 JOURNAL OF VASCULAR SURGERY Volume 43, Number 6 Hoornweg et al 1097 management of 28 patients with acute traumatic thoracic include the abdomen and both groins in the operative field, ruptures and to evaluate whether endovascular treatment is thus providing access to both common femoral arteries and a valuable alternative to open surgical repair in the emer-allowingency setting. mon femoral artery was exposed surgically, and vascular retroperitoneal iliac access if needed. One com- access was obtained. The endoprosthesis was delivered over PATIENTS AND METHODS a super-stiff Backup Meier guidewire (Boston Scientific/ We retrospectively reviewed all patients with a traumatic thoracic aortic rupture treated with an endovascular oroscopic guidance within the descending aorta for deploy- Schneider, Bülach, Switzerland) and positioned under flu- graft between June 2000 and April 2004 in four level 1ment. A 7F or 5F sheath was introduced percutaneously trauma centers. All patients sustained a violent traumatic through the contralateral femoral artery, and intraoperative injury. In 27 patients, this involved sudden deceleration arteriography was performed by using a mobile C-arm (24 traffic accidents and 3 falls from a great height), and (OEC or OEC 9600; OEC Medical Systems, Salt patient was crushed by a fork-lift truck. All patients werelake City, Utah) and a 5F pigtail catheter (Alfa Flow treated with a thoracic endovascular graft. The following [Optimed, Ettlingen, Germany or Cordis, Waterloo, Bel- After deployment, further expansion of the endo- parameters were studied: age, sex, injury severity score, typegium]). of endovascular graft, endovascular operation time, length vascular graft was performed by ballooning at the proximal of stay in the hospital, length of stay in the intensive care and distal anchoring zones as needed. Systemic heparin was unit (ICU), and mortality. In addition, we recorded follow-contraindicateup data consisting of computed tomographic angiography dominal, or bone injuries and was therefore never used. At as a result of accompanying cerebral, ab- (CTA) and plain chest radiographs at regular intervals. the end of the procedure, a final angiogram was made to In all our trauma patients, initial trauma care was ac-confircording the Advanced Trauma Life Support guidelines. endoleaks. the position of the graft and to detect possible 23 Hence, plain chest radiographs were made, as well as pelvic A CTA and was made on the first postoperative day. All and cervical spine radiographs. In all but one patient, thepatients received a platelet inhibitor (aspirin 100 mg daily) presence of aortic injury was immediately highly suggestive for 3 months unless this was contraindicated. Follow-up on plain chest radiographs. Twenty-six patients showed a consisted of CTA and four-directional plain chest radio- after 6 and 12 months and yearly thereafter. Only widened mediastinum, and one patient had a fracture of thegraphs second rib. In concordance with the Advanced Trauma Lifefour-directional chest radiographs were made if the aorta Support guidelines, diagnosis was confirmed with contrastenhanced CTA. We used a spiral computed tomographic as mean and standard deviation or median and interquartile showed complete healing on CTA. The data are presented scanner (Mx8000 [Philips Medical Systems, Eindhoven, range (IQR) and were analyzed by using SPSS Version 11.5 The Netherlands], a Somatom 40 or 64 Sensation scanner(spss Inc, Chicago, Ill). [Siemens, Erlangen, Germany], or Aquilon Multi [Toshiba, Nasu, Japan]). For intravenous contrast enhancement, we RESULTS used power injection of 100 or 110 ml of Omnipaque 300 or We treated 28 patients with a traumatic thoracic aortic Omnipaque 350 (Nycomed Inc, Princeton, NJ) at a rate ofrupture with thoracic endografts in four academic centers: 3 ml/s. Scanning was initiated 30 seconds after the start ofacademic Medical Center, Amsterdam (13 patients); Uni- Medical Center, Utrecht (6 patients); University contrast injection by using a 5-mm slice thickness and aversity 6.25 mm/s table incrementation (pitch, 1.250) or a 2-mmMedical Center, Groningen (4 patients); and University slice thickness and table movement of 11 mm per rotationhospital Maastricht (5 patients). All participating centers (pitch, 5.5). Sizing of the endovascular graft was basedare level 1 trauma centers and were experienced in endo- procedures. Of the 28 patients, 19 (68%) were on CTA measurements. The length of the graft dependedvascular on available sizes from the manufacturer and was chosen asprimary trauma admissions, whereas the other patients were short as possible. Graft diameter was determined by usingtransferred from level 2 centers either because of the severity of their injuries or because there was suspicion of a curved linear reconstructions to determine the central lumen line 24 or by measurement of the smallest diameter on traumatic aortic rupture. the oblique aortic cross sections on CTA. During the study period, two additional patients admitted at the participating centers underwent conventional After initial assessment and stabilization, patients were transferred to the operating theater. Acute life-threatening thoracic aortic surgery for repair of an acute rupture. One injuries such as intracranial or intra-abdominal hemorrhage patient underwent conventional repair for logistic reasons were treated first, followed immediately by the endovascular procedure. A vascular surgeon and interventional radi- perform an endovascular procedure. In the second patient, because there was no experienced surgeon available to ologist performed the endovascular procedure by using two it seemed on CTA that the rupture was located between the types of endografts: the Gore thoracic aortic graft (TAG; left carotid and left subclavian artery, and, therefore, an W. L. Gore & Associates, Flagstaff, Ariz) and the Talentendograft would have covered the carotid artery. This patient thoracic endoprosthesis (Medtronic AVE, Santa Rosa, died during the open surgical procedure. No patient died Calif). Patients were positioned in the supine position andbefore an intended endovascular repair, and no patient was treated under general anesthesia. Drapes were arranged to denied treatment because of a dismal prognosis.

3 1098 Hoornweg et al JOURNAL OF VASCULAR SURGERY June 2006 Table I. Concomitant injuries Table II. Operation characteristics (n 28) Concomitant injury n % Parameter Data Head injuries Closed head injury Skull fracture Facial fracture Abdominal Spleen Liver Renal Diaphragm Pancreatic contusion Mesenterial hematoma Cardiopulmonary Pulmonary contusion Hemothorax Pneumothorax Fractures Rib Upper extremity Pelvic Lower extremity Vertebral Skin Burns on 20% of TBSA TBSA, Total body-surface area. All concomitant injuries are listed in Table I. The median time between hospital arrival and endovascular treatment was 3 hours 42 minutes (range, 30 minutes to 4 days 22 hours 15 minutes). Twenty-two patients were treated within 24 hours of the original trauma, two patients were treated the next day, one patient was treated after 3 days, and another patient was treated after 5 days. In seven patients, the grafts had to be brought into the hospitals, with a maximum delay of 2 hours, whereas the other 21 patients were treated with grafts that were readily available in the hospitals. In two patients we noted an extreme delay. In one patient (with a delay of 4 days 22 hours) this was due to disagreement among vascular surgeons and cardiothoracic surgeons on the indication for endovascular treatment. This was the first endovascular case for that specific center. In the other patient (with a delay of 2 days 23 hours), this was simply because the diagnosis was not recognized on the initial CTA, and the plain chest radiographs of this patient showed no indication of a traumatic aortic rupture. We used the Gore TAG endovascular graft in 11 patients and the Talent thoracic endovascular graft in 17 patients. In one patient, we preoperatively planned to place two endovascular grafts to optimize proximal and distal seal. In all other patients, one graft was placed. The median diameter of the endovascular graft was 28 mm (IQR, mm), and the median length was 102 mm (IQR, mm) (Table II). Initial endovascular graft placement was successful in 25 (89%) of 28 patients. One patient showed a proximal type I endoleak on the intraoperative completion angiography, and it was successfully treated by ballooning the proximal part of the endovascular graft. In a second Median endovascular operating time, min (IQR) 58 (47-88) Left subclavian artery covered 9 (32.1%) Stent type Gore 11 (39.3%) Talent 17 (60.7%) Median diameter aorta, mm (range)* 25 (17-36) Median stent diameter, mm (IQR) 28 (28-34) Median stent length, mm (IQR) 102 ( ) Transfemoral access 100% Median length of stay in ICU, d (IQR) 9 (2-11) Median length of stay in hospital, d (IQR) 24 (14-39) IQR, Interquartile range; ICU, intensive care unit. Table III. Hospital mortality Patient no. Cause of death 1 Diffuse intra-abdominal and pulmonary bleeding with multiple organ failure 2 Severity of the accompanying injuries and an incurable psychiatric illness 3 Cervical spinal cord lesion, high-grade burns, inhalation trauma, and pneumonia 4 Tracheobronchial edema and subsequent airway obstruction after detubation Time after intervention patient, the common carotid artery was initially covered by the endovascular graft. This was resolved by pulling back the endovascular graft with an inflated balloon. In a third patient, a Talent graft with a diameter of 28 mm and length of 102 mm was chosen according to preoperative measurements. After deployment, caudal migration of a few centimeters of the stent graft was noticed. A second endovascular Talent graft with a diameter of 34 mm and length of 100 mm was successfully placed more proximally. There was no conversion to open surgery, and the secondary technical success rate was 100%. In nine patients, the rupture was located near the origin of the left subclavian artery. To obtain adequate proximal seal in these patients, intentional covering of the left subclavian artery was necessary. None of these patients experienced upper extremity ischemia or neurologic problems. In all cases, a transfemoral approach was successful. One patient required intraoperative repair of the common femoral artery because of an iatrogenic dissection. There were no postoperative complications related to the transfemoral access. Although no systemic heparin was used, no thromboembolic complications were found. Blood loss during the procedure was minimal and mainly caused by leakage through the valves of the large sheaths used during graft insertion. In 8 patients, 10 emergency procedures were performed before the endovascular procedure: 3 surgical 1h 5d 6d 26 d

4 JOURNAL OF VASCULAR SURGERY Volume 43, Number 6 Hoornweg et al 1099 Table IV. Recent case series of acute endovascular repair Study n Technical success Stent-related mortality Comorbid mortality (%) Follow-up (mo) Time until repair Paraplegia Rousseau et al (2005) 8 100% 0 0 (0%) NR* 14 d 0 Neuhauser et al (2004) % 0 0 (0%) 26.4 (6-68) 24 h 0 Wellons et al (2004) 9 100% 0 1 (11.1%) NR ND 0 Amabile et al (2004) 9 100% 0 0 (0%) 51.1 (3-41) 24 hto 21 d 0 Dunham et al (2004) % 0 1 (6.3%) 10.7 (3-30) 24 hto 25 d 0 Melnitchouk et al (2004) % 0 2 (13.4%) 31.9 (0-60) Median 2 d 0 Karmy-Jones et al (2003) % 0 3 (27.3%) h 0 Marty-Ane et al (2003) 9 100% 0 0 (0%) h to 5 d 0 Daenen et al (2003) 7 100% 0 1 (14.3%) 9 (1-18) A few hours to 17 mo 0 Orend et al (2002) % 0 1 (9.1%) 14 (1-26) 3-36 h 0 Czermak et al (2002) % 0 0 (0%) 19.5 (2-38) ND 0 Lachat et al (2002) % 1 0 (0%) hto 3wk 0 Thompson et al (2002) 5 100% 0 0 (0%) h to 5 d 0 This series % 0 4 (14.3%) 25.6 (1-56) 0 ND, Not defined. *Not reported: only the mean follow-up of emergency and delayed procedures together of 46 months (13-90 months) was reported. Not reported: in the discussion, a maximum of 12 months was reported. repairs of an open fracture, 6 laparotomies, and 1 thoracot-patientomy. The median operating time for the endovascular trauma patients with an acute thoracic aortic rupture re- with multiple injuries. Clearly, management of procedure was 58 minutes (IQR, minutes). mains a challenge. The overall hospital mortality was 14.3% (n 4), and In the last decades, various surgical strategies have been all deaths were unrelated to the aortic rupture or graftexplored to improve survival and morbidity in these com- cases. Jahromi et al summarized and compared several 6 placement (Table III). The median length of stay in theplex ICU was 9 days (IQR, 2-11 days), and the median total techniques in a systematic review. Mortality rates were 15% length of stay in the hospital was 24 days (IQR, 14-39when surgery was performed with aortic cross clamping and days). There were no patients with neurologic complications. when a Gott shunt was used, 17% when a left heart bypass no use of adjuncts to preserve distal aortic perfusion, 8% The median follow-up for survivors was 26.5 months was used, and 10% when a partial cardiopulmonary bypass (IQR, months). During follow-up, one patient was used. The paraplegia rate was 7% for repair with aortic died of unrelated causes 1 year after the initial trauma. One cross clamping, 4% with a Gott shunt, 0% with left heart graft-related complication was detected during routine bypass, and 2% with cardiopulmonary bypass. Many others 5,26-32 follow-up at 3 months. The four-directional chest radiographs showed a partial collapse of the proximal part of the Besides the surgical technique, optimal timing of the have confirmed these results. Gore TAG prosthesis, and CTA showed a pseudoaneurysm surgical procedure is still a subject of debate. Because of the at the site of the original aortic rupture. The patient waserious comorbidities seen in these patients, aortic repair asymptomatic. A second endovascular graft (Talent; diameter, 28 mm; length, 166 mm) was successfully placedthe open surgery results in unacceptably high risks for these often has to be postponed because the invasive character of inside the TAG, and the patient was discharged the samepatients. 33,34 Some studies have shown that mortality of day. Data concerning this patient were previously described the open aortic repair can be decreased when deliberately 25 10,35-37 in a case report. There were no other late endovascular delayed surgical repair is performed. Still, even with graft-related complications, such as endoleaks or graft migration. One patient was lost to follow-up after 3 months. of the patients experience a total aortic rupture, mostly careful blood pressure monitoring, approximately 2% to 5% 4,9,10 within 1 week after the trauma. DISCUSSION Endovascular repair is a minimally invasive procedure Patients undergoing thoracotomy for repair of acute that invalidates all of these reasons for postponing repair of thoracic aortic rupture show significant mortality and morbidity rates. Paraplegia is a especially feared complication, for endovascular grafting because there is no absolute need traumatic aortic rupture. Additional injuries are no obstacle with remarkably high incidences ranging from 2.3% to for systemic heparinization and because the procedure causes 5, %, depending on the chosen operative technique. minimal blood loss. After treatment of acute life-threatening Additionally, because patients with a traumatic aortic rupture already have lung contusions, thoracotomy often loss), the endovascular graft placement can be performed in injuries (intracranial hemorrhage or intra-abdominal blood (24%-65%) results in prolonged respiratory insufficiency the same session. This study revealed a median operating and infectious complications. 6,7,27 Furthermore, open surgery requires heparinization, which is disadvantageous for which is favorable compared to results in the time of less than 1 hour (58 minutes; IQR, minutes), litera-

5 1100 Hoornweg et al JOURNAL OF VASCULAR SURGERY June 2006 ture. 14,18,22 Also, the median ICU stay (9 days; IQR, 2-11thermore, maximal aortic cuffs lengths are 34 mm for days) and total hospital stay (24 days; IQR, days) Zenith, 30 mm for Medtronic, and 45 mm for Gore, so were shorter when compared with stays reported for openmultiple cuffs may be required to obtain an adequate seal surgery, 4,38 but these two parameters are more dependent and sufficient length of the stent-graft. This results in on accompanying injuries than on the minimally invasive significant risk for disconnection of the multiple aortic vascular intervention. cuffs. In contrast, the Endologix Powerlink (Endologix, Table IV summarizes the early results of endovascular Irvine, CA) aortic cuffs with lengths of 5.5 or 7.5 cm are repair for traumatic thoracic aortic ruptures in 13 recentnot much different from the aortic grafts we have used in studies. Unfortunately, the interval between trauma and this study. Probably in the future dedicated endovascular endovascular repair varies widely among these series, and grafts will be designed for this specific group of patients this makes it difficult to make comparisons. Accompanying with traumatic ruptures of the thoracic aorta. injuries seem to be the primary cause of mortality in patients treated with an endovascular graft. Technical success cular grafting procedures. In our series, one patient had a Endoleak is a unique problem associated with endovas- for endovascular procedures is excellent, with a range oftype I endoleak, which could immediately be resolved. 81.5% to 100%, and endovascular graft-related mortality is Type I endoleaks can usually be repaired either by additional only 1% ,18-22, ,21,39,40 The reported incidence of paraple - balloon dilation or placement of a second graft. gia after endovascular repair for traumatic thoracic ruptures Fattori et al and Melnitchouk et al both reported one is 0%, which is very favorable when compared with the paraplegia rate associated with open repair (2.3%-25.5%). The respectively spontaneous thrombosed on follow-up or was patient with a type III endoleak on follow-up. endoleak T h e 5,26 relatively short lengths of the endovascular grafts and thesuccessfully treated with a second endovascular graft. However, Lachat et al described a patient with a complete 18 absence of the need for aortic cross clamping and, thus, compromising the collateral blood flow to the spinal cordrupture within 12 hours of surgery. This was probably 42 could explain these good results. caused by an undetected incomplete proximal seal. This Our study also did not reveal complications after theindicates the risk of endoleaks and the need for accurate left subclavian artery was covered during repair of traumaticdetection. thoracic aortic ruptures, and this is equivalent to other Another point of concern is long-term follow-up. reports. 12,20,43,44 Although coverage of the left subclavian The patient population with traumatic aortic ruptures is artery is generally well tolerated and does not cause com-relativelplications, cerebral problems and upper extremity claudica- and exceeds the current experience with endovascular young, and their life expectancy is considerable 21 tion have been described. Obviously, the presence of a grafts. Consequently, patients with an endovascular graft left-sided mammary arterial bypass and a dominant left for a traumatic aortic rupture will have to be closely vertebral artery are contraindications for covering the left monitored for a long time. Currently, we perform yearly subclavian artery. CTA and conventional four-directional chest radio- for several years or until the aortic wall shows For sizing of the endovascular graft, we measured thegraphs aortic diameter between the left subclavian and carotid complete healing. From that time on, yearly follow-up artery origins and oversized the graft by 10% to 20% if can probably be reduced to yearly chest radiographs to possible. However, the aorta of patients with a traumaticevaluate the graft position and integrity of the stents. aortic rupture is quite different from the atherosclerotic Regardless, in our opinion, endovascular grafting of the diseased aortas for which these endovascular grafts were traumatic thoracic aortic ruptures is a definite treatment originally designed. Because of the available endovascular and not a bridging procedure. Some patients might graft diameters, this may result in a relatively large mis-neematch between the diameter of the aorta and the en-probably be of an endovascular nature. a second procedure, but those procedures will dografts, thus leading to extensive oversizing of the graft 40 (up to 50%). Wellons et aldescribe the use of aortic extension cuffs with a mean stent-graft diameter of 25 mm CONCLUSION (range, mm) in the treatment of traumatic aortic The considerable mortality and morbidity associated rupture. The advantage of this approach is the more appro-witpriate diameter of these cuffs. However, this advantage isconventional open surgery necessitate a new approach. On repair of traumatic ruptures of the thoracic aorta by counterbalanced by the fact that delivery systems for thesethe basis of our experience, we conclude that immediate cuffs lack sufficient length because the aortic cuffs wereendovascular repair is a safe and feasible procedure for such originally designed for infrarenal use. The AneuRx repairs, and it therefore seems to be preferable over conventional open repair. The endovascular procedure is min- (Medtronic AVE, Santa Rosa, CA) and Zenith (William Cook Europe, Bjaeverskov, Denmark) delivery devices and imally invasive; thus, delay in treatment as a result of the Excluder (W.L. Gore & Associates, Flagstaff, AZ) comorbid injury is no longer necessary. Furthermore, para- risk is reduced. However, close and long-term fol- delivery device allow only for, respectively, 55 and 61 cm toplegia reach the left subclavian artery. Assuming that only patientslow-up of the relatively young patients with a traumatic less than 6 feet tall can be treated, only a part of the thoracic aortic rupture and endovascular repair is war- European population can be treated with aortic cuffs. Fur-ranted.

6 JOURNAL OF VASCULAR SURGERY Volume 43, Number 6 Hoornweg et al 1101 AUTHOR CONTRIBUTIONS Conception and design: LLH, HJMV, JCG, MKD, RB Analysis and interpretation: LLH, MKD, HJMV, RB Data collection: LLH, MKD, JCG, JAR, HJMV, ELV, GWHS, RB Writing the article: LLH, RB Critical revision of the article: LLH, MKD, JCG, JAR, HJMV, ELV, GWHS, RB Final approval of the article: LLH, RB, MKD, JCG, JAR, HJMV, ELV, GWHS Overall responsibility: RB REFERENCES 1. Mattox KL, Wall MJJ, LeMaire SA. Injury to the thoracic great vessels. In: Mattox KL, Feliciano DV, Moore EE, editors. Trauma. 4th ed. New York: McGraw-Hill; p Richens D, Field M, Neale M, Oakley C. The mechanism of injury in blunt traumatic rupture of the aorta. Eur J Cardiothorac Surg 2002;21: Richens D, Kotidis K, Neale M, Oakley C, Fails A. Rupture of the aorta following road traffic accidents in the United Kingdom The results of the co-operative crash injury study. Eur J Cardiothorac Surg 2003;23: Fabian TC, Richardson JD, Croce MA, Smith JS Jr, Rodman G Jr, Kearney PA, et al. Prospective study of blunt aortic injury: Multicenter Trial of the American Association for the Surgery of Trauma. J Trauma 1997;42: Attar S, Cardarelli MG, Downing SW, Rodriguez A, Wallace DC, West RS, et al. Traumatic aortic rupture: recent outcome with regard to neurologic deficit. Ann Thorac Surg 1999;67: Jahromi AS, Kazemi K, Safar HA, Doobay B, Cina CS. Traumatic rupture of the thoracic aorta: cohort study and systematic review. J Vasc Surg 2001;34: Razzouk AJ, Gundry SR, Wang N, del Rio MJ, Varnell D, Bailey LL. Repair of traumatic aortic rupture: a 25-year experience. Arch Surg 2000;135: Tatou E, Steinmetz E, Jazayeri S, Benhamiche B, Brenot R, David M. Surgical outcome of traumatic rupture of the thoracic aorta. Ann Thorac Surg 2000;69: Pate JW, Fabian TC, Walker W. Traumatic rupture of the aortic isthmus: an emergency? World J Surg 1995;19: Pierangeli A, Turinetto B, Galli R, Caldarera L, Fattori R, Gavelli G. Delayed treatment of isthmic aortic rupture. Cardiovasc Surg 2000;8: Volodos NL, Karpovich IP, Troyan VI, Kalashnikova Y, Shekhanin VE, Ternyuk NE, et al. Clinical experience of the use of self-fixing synthetic prostheses for remote endoprosthetics of the thoracic and the abdominal aorta and iliac arteries through the femoral artery and as intraoperative endoprosthesis for aorta reconstruction. Vasa Suppl 1991;33: Amabile P, Collart F, Gariboldi V, Rollet G, Bartoli JM, Piquet P. Surgical versus endovascular treatment of traumatic thoracic aortic rupture. J Vasc Surg 2004;40: Czermak BV, Waldenberger P, Perkmann R, Rieger M, Steingruber IE, Mallouhi A, et al. Placement of endovascular stent-grafts for emergency treatment of acute disease of the descending thoracic aorta. AJR Am J Roentgenol 2002;179: Daenen G, Maleux G, Daenens K, Fourneau I, Nevelsteen A. Thoracic aorta endoprosthesis: the final countdown for open surgery after traumatic aortic rupture? Ann Vasc Surg 2003;17: Dunham MB, Zygun D, Petrasek P, Kortbeek JB, Karmy-Jones R, Moore RD. Endovascular stent grafts for acute blunt aortic injury. J Trauma 2004;56: Karmy-Jones R, Hoffer E, Meissner MH, Nicholls S, Mattos M. Endovascular stent grafts and aortic rupture: a case series. J Trauma 2003; 55: Kato N, Dake MD, Miller DC, Semba CP, Mitchell RS, Razavi MK, et al. Traumatic thoracic aortic aneurysm: treatment with endovascular stent-grafts. Radiology 1997;205: Lachat M, Pfammatter T, Witzke H, Bernard E, Wolfensberger U, Kunzli A, et al. Acute traumatic aortic rupture: early stent-graft repair. Eur J Cardiothorac Surg 2002;21: Marty-Ane CH, Berthet JP, Branchereau P, Mary H, Alric P. Endovascular repair for acute traumatic rupture of the thoracic aorta. Ann Thorac Surg 2003;75: Melnitchouk S, Pfammatter T, Kadner A, Dave H, Witzke H, Trentz O, et al. Emergency stent-graft placement for hemorrhage control in acute thoracic aortic rupture. Eur J Cardiothorac Surg 2004;25: Orend KH, Pamler R, Kapfer X, Liewald F, Gorich J, Sunder-Plassmann L. Endovascular repair of traumatic descending aortic transection. J Endovasc Ther 2002;9: Thompson CS, Rodriguez JA, Ramaiah VG, DiMugno L, Shafique S, Olsen D, et al. Acute traumatic rupture of the thoracic aorta treated with endoluminal stent grafts. J Trauma 2002;52: American College of Surgeons. Advanced Trauma Life Support Course For Physicians. 7th ed. Chicago: Committee on Trauma, American College of Surgeons Balm R, Eikelboom BC, van Leeuwen MS, Noordzij J. Spiral CTangiography of the aorta. Eur J Vasc Surg 1994;8: Idu MM, Reekers JA, Balm R, Ponsen KJ, de Mol BA, Legemate DA. Collapse of a stent-graft following treatment of a traumatic thoracic aortic rupture. J Endovasc Ther 2005;12: von Oppell UO, Dunne TT, De Groot MK, Zilla P. Traumatic aortic rupture: twenty-year metaanalysis of mortality and risk of paraplegia. Ann Thorac Surg 1994;58: Gammie JS, Shah AS, Hattler BG, Kormos RL, Peitzman AB, Griffith BP, et al. Traumatic aortic rupture: diagnosis and management. Ann Thorac Surg 1998;66: Hochheiser GM, Clark DE, Morton JR. Operative technique, paraplegia, and mortality after blunt traumatic aortic injury. Arch Surg 2002; 137: Kodali S, Jamieson WR, Leia-Stephens M, Miyagishima RT, Janusz MT, Tyers GF. Traumatic rupture of the thoracic aorta. A 20-year review: Circulation 1991;84(5 Suppl):III Moore EE, Burch JM, Moore JB. Repair of the torn descending thoracic aorta using the centrifugal pump for partial left heart bypass. Ann Surg 2004;240: Szwerc MF, Benckart DH, Lin JC, Johnnides CG, Magovern JA, Magovern GJ Jr, et al. Recent clinical experience with left heart bypass using a centrifugal pump for repair of traumatic aortic transection. Ann Surg 1999;230: Zeiger MA, Clark DE, Morton JR. Reappraisal of surgical treatment of traumatic transection of the thoracic aorta. J Cardiovasc Surg (Torino) 1990;31: Chiesa R, de Moura MR, Lucci C, Castellano R, Civilini E, Melissano G, et al. Traumatic rupture of the thoracic aorta. Acta Chir Belg 2003;103: Galli R, Pacini D, Di Bartolomeo R, Fattori R, Turinetto B, Grillone G, et al. Surgical indications and timing of repair of traumatic ruptures of the thoracic aorta. Ann Thorac Surg 1998;65: Holmes JH, Bloch RD, Hall RA, Carter YM, Karmy-Jones RC. Natural history of traumatic rupture of the thoracic aorta managed nonoperatively: a longitudinal analysis. Ann Thorac Surg 2002;73: Kipfer B, Leupi F, Schuepbach P, Friedli D, Althaus U. Acute traumatic rupture of the thoracic aorta: immediate or delayed surgical repair? Eur J Cardiothorac Surg 1994;8: Symbas PN, Sherman AJ, Silver JM, Symbas JD, Lackey JJ. Traumatic rupture of the aorta: immediate or delayed repair? Ann Surg 2002;235: Hemmila MR, Arbabi S, Rowe SA, Brandt MM, Wang SC, Taheri PA, et al. Delayed repair for blunt thoracic aortic injury: is it really equivalent to early repair? J Trauma 2004;56: Neuhauser B, Czermak B, Jaschke W, Waldenberger P, Fraedrich G, Perkmann R. Stent-graft repair for acute traumatic thoracic aortic rupture. Am Surg 2004;70:

7 1102 Krasznai and Blankensteijn JOURNAL OF VASCULAR SURGERY June Wellons ED, Milner R, Solis M, Levitt A, Rosenthal D. Stent-graft repair of traumatic thoracic aortic disruptions. J Vasc Surg 2004;40: Rousseau H, Dambrin C, Marcheix B, Richeux L, Mazerolles M, Cron C, et al. Acute traumatic aortic rupture: a comparison of surgical and stent-graft repair. J Thorac Cardiovasc Surg 2005;129: Balm R. Acute traumatic aortic ruptures. In: Greenhalgh RM, editor. Vascular and endovascular challenges. London: Biba Publishing; p Gorich J, Asquan Y, Seifarth H, Kramer S, Kapfer X, Orend KH, et al. Initial experience with intentional stent-graft coverage of the subclavian artery during endovascular thoracic aortic repairs. J Endovasc Ther 2002;9(Suppl 2):II Scharrer-Pamler R, Kotsis T, Kapfer X, Gorich J, Orend KH, Sunder- Plassmann L. Complications after endovascular treatment of thoracic aortic aneurysms. J Endovasc Ther 2003;10: Fattori R, Napoli G, Lovato L, Russo V, Pacini D, Pierangeli A, et al. Indications for, timing of, and results of catheter-based treatment of traumatic injury to the aorta. AJR Am J Roentgenol 2002;179: Submitted Oct 14, 2005; accepted Jan 11, INVITED COMMENTARY Attila G. Krasznai, MD, and Jan D. Blankensteijn, MD, Nijmegen, The Netherlands The authors are to be commended for collecting one of the larger series of endovascular repair of traumatic ruptures of the thoracic aorta. At the same time, the retrospective and caseselective nature of their study population limits the validity of this report. It is not unlikely that a number of patients with arch ruptures were never considered for endovascular repair early in the study period as endovascular repair was not introduced simultaneously in the four study centers. Unfortunately, the outcomes of these patients along with the ones unsuitable for endovascular repair as a result of anatomic restrictions or hemodynamic instability were not studied. Even in a retrospective setting, the study could have been much stronger if all traumatic ruptures in The Netherlands in this period were included and not just the ones that received an endovascular graft in a few trauma centers. Open surgical repair of traumatic arch ruptures is associated with high mortality and morbidity rates. Because of frequent major comorbidities, the timing of surgical repair is still controversial. Paraplegia, respiratory insufficiency, renal failure, and major bleeding are common complications after open surgical repair. Various studies have suggested better outcomes for patients who are treated in a delayed manner. The present study on endovascular repair seems to confirm endovascular repair to be a safe and highly effective procedure with fewer intraoperative and postoperative complications, even for high-risk patients. Although short-term results are encouraging, concerns have been raised regarding stentgraft failure due to the acute angle of the aortic arch in young patients and regarding stent-graft migration. Also, the issue of durability of endovascular repair is highly relevant in this patient category. With the low incidence of traumatic aortic ruptures, randomized trials are almost impossible to conduct. Prospective, population-based studies including all patients with aortic arch ruptures are probably the best attainable level of evidence on this issue. For the moment, this large retrospective study adds to the perception that immediate or early endovascular repair of traumatic ruptures of the thoracic aorta is feasible, safe, and effective in polytrauma patients even in the face of severe accompanying injuries.