Emergency Stent-graft Repair for Thoracic Aortic Injury


Submitted on Fri, 09/05/2008 - 16:36

Patrizio Castelli, MD, Roberto Caronno, MD, Gabriele Piffaretti, MD, Matteo Tozzi, MD, Chiara Lomazzi, MD, Domenico Laganà, MD, Gianpaolo Carrafiello, MD, Salvatore Cuffari, MD

Introduction Blunt thoracic aortic injuries (BTAIs) due to a deceleration trauma are highly lethal and remain a therapeutic challenge. BTAIs represent only 10% of all thoracic vascular traumas and usually involve the isthmic portion of the aorta, but up to 80% of patients with acute rupture die at the scene of injury or before reaching the operating room.1,2 In fact, for patients who survive, prognosis is still poor, because even in a properly-monitored unit, there is a 30% mortality rate within the first six hours and up to a 50% mortality rate within the first twenty-four hours.3,4 Thus, considerable attention is paid to rapid diagnosis and treatment. Over the past decades, open surgical repair has been the standard method of treatment since the first description of traumatic injury of the aorta.1 However, BTAI is rarely a simple injury and patients have severe co-morbidities, making immediate conventional repair problematic, including thoracotomy, one-lung ventilation and anticoagulation. Early postoperative mortality is still reported to range from 7.7–28%.5-9 More recently, several reports have highlighted excellent results using an endovascular approach for chronic traumatic ruptures of the descending thoracic aorta.10,11 The main advantages of the endovascular technique include the absence of thoracotomy, aortic cross-clamping and fully intraoperative heparinization. We report our ongoing experience with emergency SG treatment for the management of patients with BTAI. Materials and Methods In the last three years, 35 patients underwent endovascular repair of descending thoracic aorta, including 7 patients (20%) with BTAI after a road-traffic accident (RTA). All were male patients with a median age of 26 years (mean: 26.8 ± 9.2; range 18–45). Glasgow Coma Score (GCS) and injury severity scores (ISS) were 8 ± 5 (range 3–15; median 6) and 42.7 ± 29 (range 4–75; median 50), respectively. Since 2001, every patient admitted to the Emergency Departement (ED) with suspected BTAI was evaluated for endovascular repair (EVAR). The management team included a senior vascular surgeon and an interventional radiologist. In addition, other personnel (e.g., anesthesiologist, anesthesiology nurses, emergency department staff) were involved as appropriate for both standard management and endovascular approach of acute thoracic aortic rupture. Our protocol involved accepted steps for acute aortic rupture, including blood samples for routine laboratory studies and cross-matched, purified red blood cells, central venous and bladder catheters, a peripheral 14-G vein access, and radial artery cannulation for invasive blood pressure (BP) control. Feasibility for emergency endovascular repair (eEVAR) was assessed, if and when cardiocirculatory stability was present. Assessment was via CT-angiography (CT-A) with 3D reconstructions, carried out with the senior surgeon and the interventional radiologist to immediately confirm the feasibility of the endovascular approach, and an anesthesiologist to control the cardiocirculatory and airway support. CT-A was used to confirm the diagnosis of acute thoracic aortic rupture, assess the feasibility of the endovascular procedure, measure the morphological dimensions, exclude alternative intra-thoracic disease, and complete trauma staging. During the CT-A, the operating room staff was notified. Acute traumatic rupture is defined as disruption of the aortic wall with blood outside the adventitia and mediastinal hematoma (contained rupture) or hemothorax documented by preoperative CT-A. Location of aortic injury involved the aortic isthmus in all patients, and originated 0.5–4 cm (mean: 1.5 ± 0.5 cm) from the origin of the left subclavian artery (LSA). We detected two kinds of lesions (Figure 1): a contained circular/semi-circumferential transection (n = 5) and a localized intimal disruption (n = 2). All lesions extended 4 to 6 cm (mean: 4.4 ± 0.7) from the origin of the LSA. Suitable morphology for EVAR also included no marked tortuosity or stenosis of the aorto-iliac arteries. Six out of 7 patients (85.7%) were treated within 48 hours from the RTA. Delay between the RTA and EVAR ranged between 2 hours and 22 days (median: 15). The long delay was mostly due to hemodynamically stable patients (3 cases) who underwent SG placement two days after surgical treatment of concomitant life-threatening lesions. In 4 cases (57%), the procedure was done on the same day as the event because of a severe hemorrhagic shock. The remaining 3 patients were treated between 24 and 72 hours from admission. Repairs were performed in the operating room with the patient under general anesthesia and endotracheal intubation. A cell-saver system and cardiopulmonary bypass were available in the event that surgical conversion was needed. Inventory of various devices is kept in the operating room to allow for such urgent cases without delay. The fluoroscopic machine (Isocentric mobile C-arm, Siemens, Munich, Germany) was positioned opposite to the first operator; patients were prepared and draped for either femoral arteriotomy or retroperitoneal iliac artery approach, and emergency left thoracotomy. We used autotransfusion (Compact-Dideco®, Modena, Italy) and a fluid protocol [5%-mannitol, dopamine (3g/kg/min) plus N-acetylcysteine (600 mg i.v.)] against the ischemic-related delivery of free-radicals and the renal damage due to the contrast medium. Every patient received a short-term antibiotic prophylaxis with vancomycin (1 gr b.i.d.). A diagnostic catheter was passed under fluoroscopic guidance to the origin of the LSA through a percutaneous left brachial artery (LBA) access. The purpose of this access was twofold; first, to assess the LSA during stent-graft (SG) positioning, and second, to enable proximal forward-flow DSA. The common femoral artery (CFA) was exposed in standard fashion for device access in 5 patients (71.4%), whereas supra-inguinal incision was made for extraperitoneal access to the larger external iliac artery (EIA) in 2 cases. Heparin (2.500 IU) was then given intravenously. A pigtail catheter was advanced through the ascending aorta to perform a preliminary aortography to identify the injury site. A 20–24 Fr sheath, which was occasionally facilitated by the application of a small amount of mineral oil, was positioned. BP was pharmacologically (b-blocker or nitroprussiate) maintained at about 80 mmHg and the SG was delivered into the thoracic aorta under fluoroscopic control. Continuous trans-esophageal echocardiography (TEE) monitoring was performed during the procedure in order to assure the SG release and early detection of type 1 endoleak. All patients were treated with commercially available standard devices (Excluder/TAG, W.L. Gore & Associates, Flagstaff, Arizona, and Talent, Medtronic AVE, Santa Rosa, California). The SG was oversized 15% to 20% on the basis of the aortic diameter on the CT scan. All repairs were completed using a single SG. The LSA was occluded in 2 cases (28.5%) and partially over-stented in a further 2 patients. Completion DSA was routinely performed at the end of the procedure to confirm the adequate position of the SG, the complete exclusion of the injury and detect potential endoleaks (Figure 2). No further anticoagulation was provided in the postoperative period. All patients were admitted to the intensive care unit (ICU) immediately after the procedure. Co-existing injuries and adjunctive operative procedures and intervention are presented in Table 1. Follow-up CT-A and chest X-ray were scheduled at 1, 4 and 12 months after the procedure and yearly thereafter (Figure 3). Results Immediate technical success was achieved in all patients. SG deployment was successful in all 7 patients, and the aortic tear was successfully sealed in all cases. Mean operative time was 58.5 ± 15.2 minutes (range 45–90) and the endovascular procedure time was 48 minutes. Angiographic exposure time averaged 23 minutes. Every procedure required a minimum of 120 cc of iodate enhancement. The Excluder/TAG device was used in 5 patients (71.4%) and the Talent device in 2 cases. Mean SG diameter was 31 ± 4 mm (range 28–36 mm). The length of the covered thoracic aorta averaged 11 ± 2 cm (range 10–15 cm). Median procedure-related blood loss was 100 ml (range 50–120) and no patient required blood transfusion in relation to the endovascular procedure. ICU stay ranged from 4–56 days (mean 39 ± 36, median 28) and every patient was directly discharged to the outpatient rehabilitation clinic. Conversion to left thoracotomy or carotid-subclavian artery bypass was never required. There were no procedure-related neurological complications (paraplegia or stroke). We observed 2 major unrelated device complications: bilateral pneumonia and transitory liver impairment following abdominal packing for concomitant liver fractures. Further major or minor complications were not observed, including bleeding, paraplegia, embolization, renal or respiratory failure, or myocardial infarction. The in-hospital survival rate was 100%. Mean follow-up period was 18.7 ± 14.8 months (range 4–42). CT-A surveillance confirmed the complete exclusion of the injuries in all cases and there were no signs of device complication (migration, thrombosis, kinking, twisting, fracture). Re-intervention was never required. Discussion Traumatic thoracic aortic disruption is often associated with multiple systemic injuries; there are several aspects of open aortic repair that can be affected by, or have an effect on, coexisting non-aortic injuries, rendering the operative risk prohibitively high.7-9 These can include lateral patient positioning in patients with spinal or spinal cord injuries, thoracotomy and single-lung ventilation in the presence of significant pulmonary contusions, high-level systemic heparinization with closed head injuries, solid organ abdominal injury, or major fractures.2,11-13 Shorter operative time could produce less physiologic derangement and hypothermia, allowing definitive care while still observing damage control surgical principles, and contributing to overall improved outcomes, the main advantages of EVAR. Therefore, larger series showed a clear trend toward increased complications such as acute respiratory distress syndrome, bleeding, and sepsis with open BTAI repair, as compared with EVAR.7-9,14 In addition, EVAR repair has decreased transfusion rates of all blood products compared with open repairs.14 Decreased transfusion rates lead to a decreased coagulopathy and metabolic abnormalities that can occur with high-volume transfusions.7,9 EVAR seems also to limit cardio-circulatory and respiratory failure, and fatal thromboembolism. It probably depends on the rapidity of the procedure, the less important surgical trauma, and the reduced alterations of the hemodynamic state of the patient. The advantages are probably related to the rapidity of the procedure and less extended surgical trauma. Our experience is consistent with other series, reporting no mortality and major complications despite approaching BTAI in an acute setting, also confirming the immediate efficacy and decreased invasiveness of the endovascular approach. Apart from surgical mortality, one of the most devastating complication occurring after thoracic aortic repair remains paraplegia.1,2,4,5,13,14 Through the decades, several approaches (such as deep hypothermia and circulatory arrest, intercostal artery reimplantation, cerebrospinal fluid drainage, and distal circulatory support) have been attempted to delete the occurrence of this complication.1,4,5,13 When the interruption of the blood supply from the intercostal arteries is done, the critical factor of this operation is the duration of the aortic cross-clamp time. A spinal cord ischemic time longer than 30 minutes has been considered a critical event and an independent risk factor, regardless of the spinal cord protective techniques employed.15 Since the reduction in cross-clamping time has been unanimously suggested as the most effective measure to prevent perioperative paraplegia, although an endovascular approach is also not immune to paraplegia, the sudden deployment of the SG could probably confirm this hypothesis, obtaining a shorter aortic cross-occlusion time and not producing a steal phenomenon in the perfusion of the spinal cord. This should not be extrapolated to mean a zero risk of paraplegia with this procedure, as SG repair of aortic pathology does not allow for re-implantation of intercostal arteries. However, more than 90% of BAI occur at or around the isthmus of the thoracic aorta, where there are few strategic intercostal arteries. These injuries are also able to be repaired using a relatively short device, thereby minimizing the number of intercostal arteries excluded from the circulation. Required recommendation for endovascular thoracic aortic repair has been a minimum proximal landing zone of 10 to 15 mm for adequate repair and SG fixation. Typically, injuries to the aorta in trauma occur most commonly at the isthmus because of the relatively fixed nature of the thoracic aorta at this point.1,7-9 This invariably is in relatively close proximity to the orifice of the LSA.16-20 These guidelines come from experience with aneurysms of the thoracic aorta, but may not correlate with the necessary landing zone for traumatic thoracic aortic disruption repair, especially in younger patients who have non-diseased aortas that may enable easier seal, better fixation and decreased chance for migration. When there is an inadequate length of normal aorta between the LSA and the aortic pathology, there are two options available to the surgeon. The first is to cover the LSA as required and adopt a wait-and-see policy regarding the status of the left arm.16,17 Some debate exists regarding whether pre-procedural carotid-subclavian bypass is required if the LSA is covered by the SG during repair.16-18 Moreover, we have found that in most instances this requires us to cover the LSA origin. Our patients did not require further intervention and did not suffer ischemic events. This further evidence demonstrates the safety of intentional coverage of the LSA, which did not have upper extremity complications due to the extensive arterial collateralization in this area. We therefore feel that intentional coverage of the LSA allows for the most optimal repair of injuries at the aortic isthmus, with minimal risk of complications, and would advocate this maneuver for injuries at this anatomic location. In addition, one of the challenges in using SG technology in the proximal thoracic aorta is the curve of the distal arch, which can be highly angulated.17 When treating BTAI, it is important to use a very flexible system. Many patients can be unstable due to acute disruption, making care difficult. Typically, hemodynamic instability could be related to other processes, such as intra-abdominal injuries or orthopedic injuries. If, however, the patient is truly unstable as a result of the aortic disruption, immediate treatment has to be provided.21-25 The significant lower rates of morbidity and mortality for stent grafting of sub-acute or chronic aortic lesions demonstrate the benefit and advantage offered by this minimally-invasive technique.10,11 With no mortality, our series suggests that this might also be valid in the acute situation of traumatic rupture of the descending aorta. In fact, the endovascular approach is suitable in an emergency setting because it allowed us to quickly stabilize the patient hemodynamically and treat the associated visceral lesions in the best cardio-circulatory condition when compared to conventional approach.7-9 However, the quality of results is determined by the time delay between injury and endovascular management.2,26,27 The mean delay in our study was 48 hours, a time at which surgical risk is decreased, given the stabilization or resolution of associated injuries. Continued technical advances and improved SG devices, as well as gained experience, suggest that endografting at the aortic isthmus may be possible at the time of initial presentation or if patients abruptly deteriorate at follow-up.21,27 Moreover, there is no need for a custom-made SG. Rupture at the aortic isthmus shows very little variation between patients with regards to location and diameter, and short segment lesion at or within 30 mm from the origin of the LSA.1,2,19,20 In addition, aortic diameters are somewhat standardized and rupture of the aortic isthmus usually occurs following motor vehicle accidents in younger patients with non-calcified aorta, which facilitates the endovascular procedure. The management suggested here with available SGs for treatment at the time of initial presentation or if early abrupt deterioration occurs is probably optimal for these patients with evolutive injury, because they are associated with increased surgical risk. The pitfall of endovascular management compared to surgery is the need for a multi-disciplinary team that must be available to perform the procedure on an emergency basis. Generally, co-morbid injuries, rather than surgical technical factors, represent the primary cause of mortality in patients with BTAI.28 Co-existing lethal injuries have been reported in up to 90% of patients with BTAI.1,2,28 In a review of 75 traumatic aortic ruptures, it has been noted that a higher ISS correlated with death at the scene and a second lethal injury occurred in 41.2% of those victims.28,29 We also experienced associated lethal injuries to be common. In addition, several papers have identified a higher ISS in non-survivors versus survivors of BTAI. A higher ISS particularly correlated with mortality and patients who arrived without vital signs, in shock, or with high ISS had worse outcomes.1-4 In our series, aside from four patients who were admitted with signs of massive bleeding and profound hemorrhagic shock (mean GCS/ISS: 3.75 and 62.5, respectively), we did not observe fatal outcome, which compares favorably to that noted in the literature review for conventional open repair.29 This does not mean that ISS is not a reliable prognostic score, but that the rapidity to restore the cardio-circulatory setup and the less-extended surgical trauma of the endovascular approach could improve the otherwise fatal outcome, even though we did not directly compare this data with a control group of patients treated with open surgery. Moreover, the rapidity of this approach could perform the adjunctive surgical procedures for the associated multiple non-lethal injuries that are also common in victims of blunt BTAI and account for more than 70% of the patients in our series, including combined chest and abdominal injuries, orthopedic injuries and neuro-trauma. Conclusion Our initial experience attests to the feasibility and potential attractive alternative surgical option for the treatment of DAA and TAAA for high-risk patients. It is less stressful than conventional repair, with encouraging primary results. Our experience shows that it can be used in the acute trauma setting, with an acceptable time between diagnosis and repair. In addition to a shorter procedure time, advantages of the eEVAR involved rapid stabilization of the patient, avoiding the need for extracorporeal circulation and aortic cross-clamping with its risk of paraplegia and feared side effects of systemic heparinization that could reduce perioperative complications. A short neck between LSA and injury is not a contraindication for SG placement. Despite the fact that half of our cases required the coverage of the LSA, no complication occurred, possibly due to the presence of collateral flow. A dedicated state-of-the-art physician team, operating room suite and a wide SG inventory are preferable. The collaboration between surgeons, radiologists, and anesthesiologists helps to resolve many difficulties, such as hemodynamic stabilization, interpretation of emergent CT scans, appropriate and quick imaging investigations to determine if aneurysm morphology and vascular access are suitable for EVAR, and the choice of an appropriate device. Concerns with regards to the long-term integrity of the material are difficult to assess given the limited experience. However, even for younger patients that could undergo standard surgical management, this promising approach may improve prognosis because the main goal is to preserve life, especially for those at major risk. lelepiffa74@libero.it