PAPERS OF THE 134TH ASA ANNUAL MEETING
Repair of Extensive Aortic Aneurysms A Single-Center Experience Using the Elephant Trunk Technique Over 20 Years Anthony L. Estrera, MD, Harleen K. Sandhu, MD, Charles C. Miller, PhD, Kristofer Charlton-Ouw, MD, Tom C. Nguyen, MD, Rana O. Afifi, MD, Ali Azizzadeh, MD, and Hazim J. Safi, MD
Objectives: We report the early and late outcomes after repair of extensive aortic aneurysms using the 2-stage elephant trunk (ET) technique. Background: Management of aneurysm involving the entire aorta is a significant challenge. Given the anatomical complexity, the staged ET procedure was devised. A paucity of long-term data of outcomes of this approach exists. Methods: A single-center retrospective analysis of a prospectively collected database of all patients undergoing repair for extensive aortic aneurysm was performed. Results: Between 1991 and 2013, we repaired 3012 aneurysms of the ascending or thoracoabdominal aorta. Of these, we performed 503 operations in 348 patients using the ET technique. Mean age was 62.4 ± 14.3 years, and 156/346 (45.1%) operations were in women; 288 patients underwent first-stage ET with 157 receiving a complete second-stage repair. Index repair early mortality was 29/317 (9.1%). Completion stage early mortality was 17/186 = 9.1%. Stroke after first-stage ET repair was 10/297 (3.4%) and immediate neurologic deficit after the second-stage ET repair was 6/206 (2.9%). In the 131 patients who did not receive a second-stage repair, 17.8% died in the interval between 31 and 45 days. Conclusions: Extensive aortic aneurysm is a complex problem, but it can be managed safely with a 2-stage open procedure. Those patients who could not complete the completion repair fared poorly. Better predictors for early outcome need to be determined. The use of ET technique remains a valuable approach for repair of extensive aortic aneurysm. Keywords: aorta, aortic aneurysm, aortic dissection, aortic surgery, elephant trunk repair, endovascular surgery (Ann Surg 2014;260:510–518)
T
he management of extensive aortic aneurysms (EAAs); aneurysms involving the entire ascending, transverse arch; and thoracoabdominal aorta still remains a therapeutic challenge and continues to evolve. Because of the anatomical course of the native aorta, anteriorly originating from the base or root of the heart and extending superiorly and then posteriorly within the thoracic cavity, access for treatment has evolved to a staged approach using a median sternotomy followed in a second stage by a thoracoabdominal incision. The difficulties with the management of these patients result first from ischemic insults that may occur to potentially all organ systems, especially the nervous system, and second from the technical aspects of preventing bleeding after repair.1 In 1983, Borst et al2 first described the elephant trunk (ET) technique for EAAs in which a segment of graft was left “dangling”
in the descending thoracic aorta after initial repair of the ascending and transverse aortic arch. Then, after a period of recovery, the secondstage ET (ET2) was then accessed via a posterior-lateral thoracotomy or thoracoabdominal incision for completion of the repair of the EAA. The utility of the “dangling” ET allowed for ease of access of the distal arch graft without manipulation of the structures of the aortopulmonary window that could potentially lead to pulmonary artery injury and extensive bleeding. Over the course of the next 3 decades, steady improvement in outcomes has occurred after EAA with a reported mortality from 3% to 11%.1,3–11 With the advent of endovascular therapies, hybrid procedures were developed and applied with the ET procedure for the management of EAA.12 The “frozen” ET (FET) procedure, in which a thoracic stent graft is deployed into the descending thoracic segment during the circulatory arrest period, was eventually devised allowing for extensive repairs to be performed without the need for the secondstage posterior open lateral thoracotomy approach.5,13 Differing techniques of the FET procedure were developed with premanufactured combined graft and stent-graft configurations.9 Furthermore, these extensive procedures have been applied to pathologies not only limited to aneurysmal disease but also to aortic dissection as well as genetically associated disorders such as Marfan syndrome.14,15 We previously examined our groups’ experience with the management of EAAs.4,16 Our general approach in the management of EAA has been to performed the symptomatic segment first regardless of the segment. In the setting of no symptoms, then we prefer repairing the ascending and arch first followed by the descending or thoracoabdominal segment as the second stage. In this work, we established the importance of completion of the second stage of the procedure because a significant risk for death from rupture existed if the second stage was delayed. The purpose of this study was to review our experience with the ET approach examining both early and late outcomes over the past 20 years.
METHODS The Committee for Protection of Human Subjects for the University of Texas Medical School at Houston, the local institutional review board, approved this study. Indications for intervention included aneurysm more than 5.0 cm, or increase in aortic diameter greater than 0.5 cm per year, and complicated acute aortic syndromes including acute aortic dissection or rupture.
Definitions From the Department of Cardiothoracic and Vascular Surgery, Clinical Science Program, The University of Texas Houston-Medical School Memorial Hermann Hospital, Houston, TX. Disclosure: The authors declare no conflicts of interest. Presented at the 134th Annual Meeting of the American Surgical Association, Boston, MA. Reprints: Anthony L. Estrera, MD, Department of Cardiothoracic and Vascular Surgery, The University of Texas Houston Medical School, 6400 Fannin Street, Suite 2850, Houston, TX 77030. E-mail: [email protected]. C 2014 by Lippincott Williams & Wilkins Copyright ISSN: 0003-4932/14/26003-0510 DOI: 10.1097/SLA.0000000000000892
The “first” stage ET (ET1) procedure referred to the ascending and transverse arch procedure via median sternotomy whether it was performed initially or for completion. The ET2 procedure referred to the descending or thoracoabdominal procedure approached via the thoracotomy whether it was performed initially or for completion. “Initial” referred to the procedure that was performed first chronologically with “completion” referring to the procedure that occurred after the initial procedure completing the repair. These definitions take into consideration when “reversed” ET (rET) procedures were performed. The rET referred to performing the descending thoracic
510 | www.annalsofsurgery.com
Annals of Surgery r Volume 260, Number 3, September 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
segment initially with invagination of the proximal graft for the anastomosis. Then at a later time when the transverse arch is performed, the rET can be accessed and used for the arch procedure. Prophylactic ET referred to performing the ET procedure for EAA but when the descending thoracic segment that was less than 5 cm. The ET is available for access in the event that the descending segment enlarges but was not intended to be immediately used, hence prophylactic. We considered aneurysms with dissection acute if surgery was performed within less than 14 days from the onset of pain, and chronic if it was performed after 14 days. Early mortality included in-hospital death and death occurring within 30 days of surgery. Interval mortality was defined as death that occurred between 31 days and 45 days after the index procedure. Patients with stroke, identified by a thorough neurologic examination and computerized tomography or magnetic resonance imaging of the head, were excluded from the neurologic deficit group; that is, neurologic deficit refers to paraplegia and paraparesis only. Postoperative immediate neurologic deficit was defined as paraplegia or paraparesis observed upon the patient awakening from anesthesia, regardless of severity. Delayed neurological deficit was defined as new paraplegia or paraparesis occurring postoperatively in a previously neurologically intact patient. The glomerular filtration rate (GFR) was calculated by the Cockcroft-Gault method, and definitions of other variables were described in previous reports.17,18
Surgical Approach The ET1 procedure, performed in similar fashion to standard surgery of the ascending aorta and transverse arch, has been described previously.19 Briefly, the first stage is performed in a similar fashion to standard surgery of the ascending aorta and transverse arch with the exception of the insertion of the Dacron graft into the descending thoracic segment. Previously, the Dacron graft was invaginated for 10 cm and the distal anastomosis of the proximal descending thoracic aorta distal to the left subclavian artery was performed to the folded edge of the graft. Introduction of the collared graft (Siena graft, Vascutec, Terumo, Scotland, UK) has eliminated the need for invagination of the graft and has allowed for disparate sizes of the descending and arch aorta to be managed using a single graft. In managing chronic aortic dissection, the dissection flap in the descending thoracic aorta is fenestrated to permit flow to both lumens. In the setting of acute dissection, the graft is placed into the true lumen with the intention of sealing the false lumen. The FET may be used in cases that involve acute dissection with distal mal-perfusion. The ascending aorta and the transverse arch are excised, leaving an island of the brachiocephalic arteries. In cases that the great vessels are enlarged, the premanufactured multibranched grafts are used for individual bypasses to each vessel. After completion of the distal arch reconstruction, the proximal aortic reconstruction is completed. For cerebral protection, retrograde cerebral perfusion (RCP) was used in conjunction with deep hypothermic circulatory arrest with a mean RCP time of 42 minutes (range 0–85 minutes). The ET2 procedure is carried out in similar fashion to our standard descending thoracic or thoracoabdominal aortic aneurysm repair.20 After induction of anesthesia, a 14-gauge needle is inserted in the intervertebral space between L3 and L4 and drains the cerebrospinal fluid. The chest is entered through a left thoracoabdominal incision (sixth intercostal space). The left groin may be exposed for access of the common femoral artery for distal aortic perfusion. The pericardium is opened posterior to the left phrenic nerve, and the left atrial appendage or pulmonary vein is exposed for the venous portion of left heart bypass. An arterial cannula is inserted into the common femoral artery or distal side branch graft of the aortic graft for distal aortic perfusion. Fairly recently, sterile ultrasound has been employed ultrasound to identify the end of the previously placed ET. This allows confirmation that the graft is not kinked and can be accessed expedi C 2014 Lippincott Williams & Wilkins
Outcomes of Elephant Trunk Technique
tiously when the aorta is opened because no proximal aortic clamp is used. Also recently, to decrease the cardiac output momentarily, the pericardium along the lateral ventricular wall is opened and an epicardial pacemaker wire is placed. Using rapid ventricular pacing, the systolic pressure and pulse pressure are decreased allowing for a relatively bloodless opening and access of the ET graft. The descending thoracic aorta is opened at the level of the previously identified graft and the ET is controlled and then clamped. Rapid ventricular pacing is terminated and normal hemodynamics are reestablished. The remaining of the descending thoracic or thoracoabdominal procedure is completed.
Statistical Methods Data were collected from chart reviews done by a trained nurse evaluator and were entered into a dedicated Microsoft Access database. Analysis was retrospective. Patient follow-up was obtained by direct patient contact, telephone interview, or the National Death Index and was complete for mortality and interventions in 97.9% (339/346). Data were managed under Health Insurance Portability and Accountability Act confidentiality guidelines in a Microsoft Access database with encrypted patient identifiers. Data were arrayed in multiple formats to address a variety of important questions. For questions related to outcomes for distinct episodes of care, operation was the unit of analysis, because about half of the patients had more than 1 operation. In cases where the patients had aneurysm-related symptoms, the symptomatic section was repaired first, and this resulted in roughly 20% of patients having the descending or thoracoabdominal segment repaired first. Short-term outcomes were assessed by repaired segment, regardless of the temporal order of treatment, because intraoperative procedures and postoperative events are substantially dissimilar between ascending/arch surgery and descending thoracic or thoracoabdominal surgery. These types of cases could not sensibly be combined for analysis. Univariate frequency statistics for discrete short-term outcomes were computed by contingency table methods with Fisher exact P values. Continuous variables were analyzed by unpaired t test. For long-term mortality, patient (rather than episode of care) was the unit of analysis, because each person can die only once, irrespective of the number of operations. Cohort time was initiated at the date of the first operation, and patients were followed up until death or censoring by date of last contact. A National Death Index survey was conducted in March 2014 to ascertain patient survival status. Patients were stratified into 2 treatment groups—those who received only 1 aortic operation, and those who completed repair of both aortic segments. Stratified survival analysis was conducted by the Kaplan-Meier method and multivariable analysis by Cox regression.
RESULTS Patients Between 1991 and 2013, 3012 repairs of the aortic root, ascending, transverse arch, and thoracoabdominal aorta were performed. Of these, we performed 503 operations in 346 patients using the staged ET technique. Two patients underwent single-stage repair and were excluded from analysis. Of the 346 patients, in 317 patients the index repair was performed by our group, and 29 patients underwent their index repair at an outside institution. In these 317 patients, we performed 288 ET1 and 29 ET2 procedures as the index operation. Figure 1 depicts the categorization of all the patients in the series. Mean patient age was 62.4 ± 14.3 years, and 156 (45.1%) of 346 operations were in women; 297 underwent ET1 procedure and 206 patients underwent the ET2 procedure. The overall median interval between first and second stage was 2.8 months. Interestingly, the median interval between repairs was 2.3 months among the group www.annalsofsurgery.com | 511
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
Estrera et al
FIGURE 1. Distribution of patients (348) undergoing ET repair, staging, sequence, and mortality. of patients on whom we performed both the ET1 and ET2 repairs as compared with 20.3 months in those who had their index repair performed elsewhere. In 297 ET1 operations, 138 (46.3%) patients underwent concomitant aortic valve procedures. Fifty percent of the ET2 repairs (103/206) involved either extent I or extent II thoracoabdominal aortic aneurysms with the remaining involving the descending thoracic aorta. Four patients underwent completion repair using thoracic endovascular aortic repair (TEVAR), 4/177 (2.3%). Cardiopulmonary bypass and deep hypothermic circulatory arrest were used in all 297 patients. Preoperative and operative variables for the ET1 and ET2 procedures are shown in Tables 1 and 2, respectively.
Early Mortality and Morbidity The early mortality was 8.4% (25/297) after ET1 repair and 10.2% (21/206) after ET2 repair. Among the ET1 procedures, no significant perioperative risk factors for mortality were identified. Of note, acute dissection and preoperative renal failure were not statistically significant risk factors for early mortality. Mortality in patients with normal GFR (>75 mL/min/1.73 m2 was 4.6%. Among the ET2 procedures, ruptured aneurysm was the only significant risk factor for early mortality [odds ratio (OR) = 7.95, P = 0.003]. Postoperative morbidities for both stages, ET1 and ET2, are listed in Table 3. A higher incidence of sepsis, adult respiratory distress syndrome, renal dysfunction/failure, gastrointestinal complications, and delayed neurologic deficit was observed after ET2 procedures. Respiratory complications and encephalopathy occurred more frequently after ET1 procedures. After ET2, 6.3% (13/206) patients developed either immediate neurologic deficit (2.9%, 6/206) or delayed neurologic deficit (3.4%, 7/206). Of those who developed immediate neurologic deficit, none recovered. Of those who developed delayed neurologic deficit, 4/7 (57%) recovered neurologic function. Thus, 9 (4.3%) of 206 developed permanent neurologic deficit. 512 | www.annalsofsurgery.com
Late Survival For patients who completed both ET1 and ET2 procedures, the 1-, 5-, 10-, 15-, and 20-year survivals were 82.9%, 67.7%, 41.5%, 27.7%, and 13.6%, respectively. For patients who did not undergo completion repair, the 1, 5, 10, 15, and 20-year survivals were 56.1%, 39.3%, 31.2%, 21.1%, and 21.1%, respectively. Those who did not complete repair had a significantly reduced long-term survival compared with those who completed repair, P < 0.0001. (Fig. 2) The only independent risk factor for late death was age (hazard ratio = 1.025, P > 0.0001). The only independent factor that was protective against death in patients undergoing repair was completion repair, that is, those who underwent both ET1 and ET2 (hazard ratio = 1.832, P > 0.0001).
Incomplete Repair (Nonreturners)
The patients who did not return for the ET2 procedure (n = 131) were categorized into those who refused surgery (but were intended to undergo a second-stage repair) (32, 24.4%, median followup, 57.3 months), those who were unfit (30, 29.9%, median follow-up, 2.3 months), and those in whom the ET1 was performed prophylactically (69, 57.2%, median follow-up 2 months). Those who were either unfit for repair or refused surgery demonstrated a significantly worse survival prognosis when compared with those in whom ET was performed prophylactically (P > 0.0001) (Fig. 3). Of the 288 patients surviving the initial stage, 6 (2.2%) of 288 died within the 31- to 45-day optimal repair interval, suggesting that 6-week wait is safe. For the patients who were classified as unfit, the interval mortality was 17.8%.
Aortic Size Comparison of aortic sizes of the treated segment among nonreturners showed no statistical difference between the different subgroups, that is, unfit versus refusal versus prophylactic (P = 0.709). However, the aortic size of the unrepaired segment was significantly C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
Outcomes of Elephant Trunk Technique
TABLE 1. 30-Day Mortality Risk Analysis: The ET1 Procedure Patient Variable
No. Patients, N (%)
Dead,∗ N (%)
297 (100)
25 (8.4)
75 (25.3) 74 (24.9) 74 (24.9) 74 (24.9) 140 (47.1) 157 (52.9) 17 (5.7) 61 (20.5)
5 (6.7) 5 (6.8) 6 (8.1) 9 (12.2) 14 (10.0) 11 (7.0) 3 (17.7) 8 (13.1)
75 (25.2) 74 (24.9) 74 (24.9) 74 (24.9) 28 (9.4) 119 (40.1) 73 (24.5) 80 (26.9) 245 (82.5) 73 (24.6) 281 (94.6) 138 (46.3)
8 (10.7) 7 (9.5) 6 (8.1) 4 (5.4) 4 (14.3) 9 (7.6) 6 (8.2) 7 (8.8) 20 (8.2) 8 (10.9) 23 (8.2) 10 (7.3)
75 (25.2) 74 (24.9) 76 (25.6) 72 (24.2)
7 (9.3) 5 (6.8) 9 (11.8) 4 (5.6)
75 (25.2) 76 (25.6) 73 (24.6) 73 (24.6)
7 (9.3) 4 (5.3) 6 (8.2) 8 (10.9)
Total Age,∗ yrs ≤58.47 58.47–67.01 67.01–73.36 >73.36 Females Males Emergent Preoperative renal failure Preoperative GFR∗ ≤54.66 54.66–71.82 71.82–92.25 >92.25 Acute dissection Chronic dissection Coronary Artery Disease COPD Systemic Hypertension Redo-median sternotomy Retrograde cerebral perfusion Aortic valve procedures Pump time, min∗ ≤123 123–149 149–179 >179 Clamp time,∗ min ≤65 65–81 81–104 >104
OR
CI
P
—
—
0.59
1.47
0.65–3.36
0.41
2.51 1.94
0.67–9.41 0.79–4.75
0.16 0.19
—
—
0.68
1.97 0.83 0.97 1.06 0.83 1.49 0.62 0.76
0.62–6.21 0.35–1.94 0.37–2.52 0.43–2.64 0.29–2.34 0.62–3.63 0.13–2.92 0.33–1.75
0.27 0.83 1.00 1.00 0.78 0.34 0.63 0.54
—
—
0.52
—
—
0.64
∗
Variables analyzed as quartiles. CI indicates confidence interval; COPD, chronic obstructive pulmonary disease.
smaller in the prophylactic group than in the unfit or refusal group (P < 0.0001) (Fig. 4).
DISCUSSION Although modifications to the original ET procedure have been made since its conception by Borst et al in 1983, the concept of placing a conduit into the descending thoracic aorta for later use or access was innovative and continues to be an important tool in the armamentarium of the aortic surgeon. Many of current challenges of EAA relate to multiorgan damage and in specific neurologic injury such as stroke or paraplegia. Advancements have been made regarding neurologic protection as was demonstrated in this report with a stroke incidence of 3.5% from ET1 procedure and an incidence of immediate neurologic deficit of 2.9% after ET2 procedure. Regarding ET1, despite growing preference for antegrade cerebral perfusion during circulatory arrest, we continue to espouse monitored RCP via the superior vena cava based on these continued noteworthy results.21 The principle adjunct for our approach to ET2 remains the use of cerebrospinal fluid drainage, distal aortic perfusion, and moderate hypothermia. Results from this report remain consistent with our previous reports with thoracoabdominal aortic aneurysm repair.20 Challenges still remain with postoperative renal dysfunction and the ET2 repair appears associated with higher rates of complications when compared with the ET1 repair.13,22 Hence, this supports our bias to perform the ET1 initially followed by the ET2 in the asymptomatic patient because recovery from the ET1 seems to be less complicated. Whether this is related to procedure-specific stress upon the patient or the relatively C 2014 Lippincott Williams & Wilkins
short time period between extensive procedures remains unknown. Considering this, the analysis of the time period between procedures became essential because a prolonged interval period may be associated with significant mortality as was noted in this series.16 One of the difficulties in analyzing this data set related to the definitions of the specific procedures, for example, ET1 and ET2, and the sequence of when the procedure was performed. In our previous series, we only analyzed a specific subset of ET patients, that is, the patient that we performed the ET1 and then the ET2 in sequence.16 Since that report, other variations of the ET procedure have been performed including the frozen ET, reversed ET, and hybrid ET that has led to differing sequences of the ET procedure. This report detailed all patients in our series that involved a staged ET procedure that was completed or was intended to be completed. For these reasons, the procedures were designated as ET1 and ET2, on the basis of approach and exposure, and the sequence was defined as either initial or completion. These designations aided in clarifying the less frequent occurrence of a patient undergoing the ET2 initially with an rET and later completing the ET1 with utilization of the reversed ET. At any rate, the majority of cases underwent the ET1 initially followed by the ET2 for completion. This remains our preferred sequence unless, the distal segment, that is, descending thoracic aorta, is much larger in size or associated with symptoms or rupture. Expanding on the analysis of our previous work,16 this study examined the subgroup of patients that did not complete the second repair, defined as the nonreturners. It was evident and intuitive that patients in whom the ET1 was performed prophylactically fared www.annalsofsurgery.com | 513
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
Estrera et al
TABLE 2. 30-Day Mortality Risk Analysis: The ET2 Procedure Patient Variable Total Age,∗ yrs ≤54.13 54.13–64.08 64.08–70.22 >70.22 Females Males Emergent Preoperative renal failure Preoperative GFR∗ ≤54.94 54.94–75.59 75.59–105.72 >105.72 Chronic dissection Coronary artery disease COPD Systemic Hypertension Diabetes Mellitus Preoperative stroke Rupture Redo thoracic aneurysm TAAA extent Extent 1 Extent 2 DTAA ∗ Pump time,∗ min ≤16 16–37 37–72 >72 Clamp time,∗ min ≤26 26–39 39–66 >66
No. Patients, N (%)
Mortality,∗ N (%)
206 (100)
21 (10.2)
52 (25.2) 51 (24.8) 52 (25.2) 51 (24.8) 95 (46.1) 111 (53.9) 10 (4.9) 38 (18.5)
4 (7.7) 4 (7.8) 4 (7.7) 9 (17.7) 8 (8.4) 13 (11.7) 2 (20.0) 6 (15.8)
52 (25.2) 51 (24.8) 52 (25.2) 51 (24.8) 108 (52.4) 50 (24.3) 58 (28.2) 169 (82.0) 19 (9.2) 15 (7.3) 12 (5.8) 9 (4.4)
8 (15.4) 6 (11.8) 4 (7.7) 3 (5.9) 12 (11.1) 6 (12.0) 6 (10.3) 19 (11.2) 4 (21.1) 1 (6.7) 5 (41.7) 2 (22.2)
55 (26.7) 48 (23.3) 103 (50.0)
6 (10.9) 4 (8.3) 11 (10.7)
57 (27.7) 48 (23.3) 51 (24.8) 50 (24.3)
9 (15.8) 2 (4.2) 3 (5.9) 7 (14.0)
58 (28.2) 48 (23.3) 51 (24.8) 49 (23.8)
4 (6.9) 3 (6.3) 7 (13.7) 7 (14.3)
OR
CI
P
—
—
0.25
0.69
0.27–1.75
0.49
2.33 1.91
0.46–11.77 0.69–5.31
0.27 0.23
—
—
0.38
1.24 1.28 1.02 2.22 2.67 0.61 7.95 2.68
0.49–3.07 0.47–3.50 0.38–2.78 0.49–9.96 0.79–8.95 0.08–4.89 2.26–27.92 0.51–13.81
0.82 0.59 1.00 0.38 0.11 1.00 0.003 0.23
—
—
0.89
—
—
0.13
—
—
0.38
∗
Variables analyzed as quartiles. CI indicates confidence interval; COPD, chronic obstructive pulmonary disease; DTAA, descending thoracic aortic aneurysm; TAAA, thoracoabdominal aortic aneurysm extent.
TABLE 3. Perioperative Morbidity Comparing ET1 Procedure With ET2 Procedure Patient Variable Total procedures (n = 491) Respiratory complications Postoperative dialysis Postoperative renal failure Myocardial Infarction GI complications Postoperative stroke Immediate neurological deficit Delayed neurological deficit
Ascending ± Arch, N (%)
Descending, N (%)
OR
CI
P
297 (59.1) 92 (30.9) 14 (4.7) 14 (4.7) 2 (0.7) 43 (14.5) 10 (3.4) 2 (0.7) 2 (0.7)
206 (40.9) 44 (21.4) 36 (17.5) 45 (21.8) 2 (0.9) 49 (23.8) 12 (5.8) 6 (2.9) 7 (3.4)
1.65 0.23 0.18 0.69 0.54 0.56 0.23 0.17
1.09–2.50 0.12–0.45 0.09–0.33 0.09–4.95 0.34–0.86 0.24–1.33 0.06–1.13 0.04–0.79
0.01
Repair of Extensive Aortic Aneurysms A Single-Center Experience Using the Elephant Trunk Technique Over 20 Years Anthony L. Estrera, MD, Harleen K. Sandhu, MD, Charles C. Miller, PhD, Kristofer Charlton-Ouw, MD, Tom C. Nguyen, MD, Rana O. Afifi, MD, Ali Azizzadeh, MD, and Hazim J. Safi, MD
Objectives: We report the early and late outcomes after repair of extensive aortic aneurysms using the 2-stage elephant trunk (ET) technique. Background: Management of aneurysm involving the entire aorta is a significant challenge. Given the anatomical complexity, the staged ET procedure was devised. A paucity of long-term data of outcomes of this approach exists. Methods: A single-center retrospective analysis of a prospectively collected database of all patients undergoing repair for extensive aortic aneurysm was performed. Results: Between 1991 and 2013, we repaired 3012 aneurysms of the ascending or thoracoabdominal aorta. Of these, we performed 503 operations in 348 patients using the ET technique. Mean age was 62.4 ± 14.3 years, and 156/346 (45.1%) operations were in women; 288 patients underwent first-stage ET with 157 receiving a complete second-stage repair. Index repair early mortality was 29/317 (9.1%). Completion stage early mortality was 17/186 = 9.1%. Stroke after first-stage ET repair was 10/297 (3.4%) and immediate neurologic deficit after the second-stage ET repair was 6/206 (2.9%). In the 131 patients who did not receive a second-stage repair, 17.8% died in the interval between 31 and 45 days. Conclusions: Extensive aortic aneurysm is a complex problem, but it can be managed safely with a 2-stage open procedure. Those patients who could not complete the completion repair fared poorly. Better predictors for early outcome need to be determined. The use of ET technique remains a valuable approach for repair of extensive aortic aneurysm. Keywords: aorta, aortic aneurysm, aortic dissection, aortic surgery, elephant trunk repair, endovascular surgery (Ann Surg 2014;260:510–518)
T
he management of extensive aortic aneurysms (EAAs); aneurysms involving the entire ascending, transverse arch; and thoracoabdominal aorta still remains a therapeutic challenge and continues to evolve. Because of the anatomical course of the native aorta, anteriorly originating from the base or root of the heart and extending superiorly and then posteriorly within the thoracic cavity, access for treatment has evolved to a staged approach using a median sternotomy followed in a second stage by a thoracoabdominal incision. The difficulties with the management of these patients result first from ischemic insults that may occur to potentially all organ systems, especially the nervous system, and second from the technical aspects of preventing bleeding after repair.1 In 1983, Borst et al2 first described the elephant trunk (ET) technique for EAAs in which a segment of graft was left “dangling”
in the descending thoracic aorta after initial repair of the ascending and transverse aortic arch. Then, after a period of recovery, the secondstage ET (ET2) was then accessed via a posterior-lateral thoracotomy or thoracoabdominal incision for completion of the repair of the EAA. The utility of the “dangling” ET allowed for ease of access of the distal arch graft without manipulation of the structures of the aortopulmonary window that could potentially lead to pulmonary artery injury and extensive bleeding. Over the course of the next 3 decades, steady improvement in outcomes has occurred after EAA with a reported mortality from 3% to 11%.1,3–11 With the advent of endovascular therapies, hybrid procedures were developed and applied with the ET procedure for the management of EAA.12 The “frozen” ET (FET) procedure, in which a thoracic stent graft is deployed into the descending thoracic segment during the circulatory arrest period, was eventually devised allowing for extensive repairs to be performed without the need for the secondstage posterior open lateral thoracotomy approach.5,13 Differing techniques of the FET procedure were developed with premanufactured combined graft and stent-graft configurations.9 Furthermore, these extensive procedures have been applied to pathologies not only limited to aneurysmal disease but also to aortic dissection as well as genetically associated disorders such as Marfan syndrome.14,15 We previously examined our groups’ experience with the management of EAAs.4,16 Our general approach in the management of EAA has been to performed the symptomatic segment first regardless of the segment. In the setting of no symptoms, then we prefer repairing the ascending and arch first followed by the descending or thoracoabdominal segment as the second stage. In this work, we established the importance of completion of the second stage of the procedure because a significant risk for death from rupture existed if the second stage was delayed. The purpose of this study was to review our experience with the ET approach examining both early and late outcomes over the past 20 years.
METHODS The Committee for Protection of Human Subjects for the University of Texas Medical School at Houston, the local institutional review board, approved this study. Indications for intervention included aneurysm more than 5.0 cm, or increase in aortic diameter greater than 0.5 cm per year, and complicated acute aortic syndromes including acute aortic dissection or rupture.
Definitions From the Department of Cardiothoracic and Vascular Surgery, Clinical Science Program, The University of Texas Houston-Medical School Memorial Hermann Hospital, Houston, TX. Disclosure: The authors declare no conflicts of interest. Presented at the 134th Annual Meeting of the American Surgical Association, Boston, MA. Reprints: Anthony L. Estrera, MD, Department of Cardiothoracic and Vascular Surgery, The University of Texas Houston Medical School, 6400 Fannin Street, Suite 2850, Houston, TX 77030. E-mail: [email protected]. C 2014 by Lippincott Williams & Wilkins Copyright ISSN: 0003-4932/14/26003-0510 DOI: 10.1097/SLA.0000000000000892
The “first” stage ET (ET1) procedure referred to the ascending and transverse arch procedure via median sternotomy whether it was performed initially or for completion. The ET2 procedure referred to the descending or thoracoabdominal procedure approached via the thoracotomy whether it was performed initially or for completion. “Initial” referred to the procedure that was performed first chronologically with “completion” referring to the procedure that occurred after the initial procedure completing the repair. These definitions take into consideration when “reversed” ET (rET) procedures were performed. The rET referred to performing the descending thoracic
510 | www.annalsofsurgery.com
Annals of Surgery r Volume 260, Number 3, September 2014
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
segment initially with invagination of the proximal graft for the anastomosis. Then at a later time when the transverse arch is performed, the rET can be accessed and used for the arch procedure. Prophylactic ET referred to performing the ET procedure for EAA but when the descending thoracic segment that was less than 5 cm. The ET is available for access in the event that the descending segment enlarges but was not intended to be immediately used, hence prophylactic. We considered aneurysms with dissection acute if surgery was performed within less than 14 days from the onset of pain, and chronic if it was performed after 14 days. Early mortality included in-hospital death and death occurring within 30 days of surgery. Interval mortality was defined as death that occurred between 31 days and 45 days after the index procedure. Patients with stroke, identified by a thorough neurologic examination and computerized tomography or magnetic resonance imaging of the head, were excluded from the neurologic deficit group; that is, neurologic deficit refers to paraplegia and paraparesis only. Postoperative immediate neurologic deficit was defined as paraplegia or paraparesis observed upon the patient awakening from anesthesia, regardless of severity. Delayed neurological deficit was defined as new paraplegia or paraparesis occurring postoperatively in a previously neurologically intact patient. The glomerular filtration rate (GFR) was calculated by the Cockcroft-Gault method, and definitions of other variables were described in previous reports.17,18
Surgical Approach The ET1 procedure, performed in similar fashion to standard surgery of the ascending aorta and transverse arch, has been described previously.19 Briefly, the first stage is performed in a similar fashion to standard surgery of the ascending aorta and transverse arch with the exception of the insertion of the Dacron graft into the descending thoracic segment. Previously, the Dacron graft was invaginated for 10 cm and the distal anastomosis of the proximal descending thoracic aorta distal to the left subclavian artery was performed to the folded edge of the graft. Introduction of the collared graft (Siena graft, Vascutec, Terumo, Scotland, UK) has eliminated the need for invagination of the graft and has allowed for disparate sizes of the descending and arch aorta to be managed using a single graft. In managing chronic aortic dissection, the dissection flap in the descending thoracic aorta is fenestrated to permit flow to both lumens. In the setting of acute dissection, the graft is placed into the true lumen with the intention of sealing the false lumen. The FET may be used in cases that involve acute dissection with distal mal-perfusion. The ascending aorta and the transverse arch are excised, leaving an island of the brachiocephalic arteries. In cases that the great vessels are enlarged, the premanufactured multibranched grafts are used for individual bypasses to each vessel. After completion of the distal arch reconstruction, the proximal aortic reconstruction is completed. For cerebral protection, retrograde cerebral perfusion (RCP) was used in conjunction with deep hypothermic circulatory arrest with a mean RCP time of 42 minutes (range 0–85 minutes). The ET2 procedure is carried out in similar fashion to our standard descending thoracic or thoracoabdominal aortic aneurysm repair.20 After induction of anesthesia, a 14-gauge needle is inserted in the intervertebral space between L3 and L4 and drains the cerebrospinal fluid. The chest is entered through a left thoracoabdominal incision (sixth intercostal space). The left groin may be exposed for access of the common femoral artery for distal aortic perfusion. The pericardium is opened posterior to the left phrenic nerve, and the left atrial appendage or pulmonary vein is exposed for the venous portion of left heart bypass. An arterial cannula is inserted into the common femoral artery or distal side branch graft of the aortic graft for distal aortic perfusion. Fairly recently, sterile ultrasound has been employed ultrasound to identify the end of the previously placed ET. This allows confirmation that the graft is not kinked and can be accessed expedi C 2014 Lippincott Williams & Wilkins
Outcomes of Elephant Trunk Technique
tiously when the aorta is opened because no proximal aortic clamp is used. Also recently, to decrease the cardiac output momentarily, the pericardium along the lateral ventricular wall is opened and an epicardial pacemaker wire is placed. Using rapid ventricular pacing, the systolic pressure and pulse pressure are decreased allowing for a relatively bloodless opening and access of the ET graft. The descending thoracic aorta is opened at the level of the previously identified graft and the ET is controlled and then clamped. Rapid ventricular pacing is terminated and normal hemodynamics are reestablished. The remaining of the descending thoracic or thoracoabdominal procedure is completed.
Statistical Methods Data were collected from chart reviews done by a trained nurse evaluator and were entered into a dedicated Microsoft Access database. Analysis was retrospective. Patient follow-up was obtained by direct patient contact, telephone interview, or the National Death Index and was complete for mortality and interventions in 97.9% (339/346). Data were managed under Health Insurance Portability and Accountability Act confidentiality guidelines in a Microsoft Access database with encrypted patient identifiers. Data were arrayed in multiple formats to address a variety of important questions. For questions related to outcomes for distinct episodes of care, operation was the unit of analysis, because about half of the patients had more than 1 operation. In cases where the patients had aneurysm-related symptoms, the symptomatic section was repaired first, and this resulted in roughly 20% of patients having the descending or thoracoabdominal segment repaired first. Short-term outcomes were assessed by repaired segment, regardless of the temporal order of treatment, because intraoperative procedures and postoperative events are substantially dissimilar between ascending/arch surgery and descending thoracic or thoracoabdominal surgery. These types of cases could not sensibly be combined for analysis. Univariate frequency statistics for discrete short-term outcomes were computed by contingency table methods with Fisher exact P values. Continuous variables were analyzed by unpaired t test. For long-term mortality, patient (rather than episode of care) was the unit of analysis, because each person can die only once, irrespective of the number of operations. Cohort time was initiated at the date of the first operation, and patients were followed up until death or censoring by date of last contact. A National Death Index survey was conducted in March 2014 to ascertain patient survival status. Patients were stratified into 2 treatment groups—those who received only 1 aortic operation, and those who completed repair of both aortic segments. Stratified survival analysis was conducted by the Kaplan-Meier method and multivariable analysis by Cox regression.
RESULTS Patients Between 1991 and 2013, 3012 repairs of the aortic root, ascending, transverse arch, and thoracoabdominal aorta were performed. Of these, we performed 503 operations in 346 patients using the staged ET technique. Two patients underwent single-stage repair and were excluded from analysis. Of the 346 patients, in 317 patients the index repair was performed by our group, and 29 patients underwent their index repair at an outside institution. In these 317 patients, we performed 288 ET1 and 29 ET2 procedures as the index operation. Figure 1 depicts the categorization of all the patients in the series. Mean patient age was 62.4 ± 14.3 years, and 156 (45.1%) of 346 operations were in women; 297 underwent ET1 procedure and 206 patients underwent the ET2 procedure. The overall median interval between first and second stage was 2.8 months. Interestingly, the median interval between repairs was 2.3 months among the group www.annalsofsurgery.com | 511
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
Estrera et al
FIGURE 1. Distribution of patients (348) undergoing ET repair, staging, sequence, and mortality. of patients on whom we performed both the ET1 and ET2 repairs as compared with 20.3 months in those who had their index repair performed elsewhere. In 297 ET1 operations, 138 (46.3%) patients underwent concomitant aortic valve procedures. Fifty percent of the ET2 repairs (103/206) involved either extent I or extent II thoracoabdominal aortic aneurysms with the remaining involving the descending thoracic aorta. Four patients underwent completion repair using thoracic endovascular aortic repair (TEVAR), 4/177 (2.3%). Cardiopulmonary bypass and deep hypothermic circulatory arrest were used in all 297 patients. Preoperative and operative variables for the ET1 and ET2 procedures are shown in Tables 1 and 2, respectively.
Early Mortality and Morbidity The early mortality was 8.4% (25/297) after ET1 repair and 10.2% (21/206) after ET2 repair. Among the ET1 procedures, no significant perioperative risk factors for mortality were identified. Of note, acute dissection and preoperative renal failure were not statistically significant risk factors for early mortality. Mortality in patients with normal GFR (>75 mL/min/1.73 m2 was 4.6%. Among the ET2 procedures, ruptured aneurysm was the only significant risk factor for early mortality [odds ratio (OR) = 7.95, P = 0.003]. Postoperative morbidities for both stages, ET1 and ET2, are listed in Table 3. A higher incidence of sepsis, adult respiratory distress syndrome, renal dysfunction/failure, gastrointestinal complications, and delayed neurologic deficit was observed after ET2 procedures. Respiratory complications and encephalopathy occurred more frequently after ET1 procedures. After ET2, 6.3% (13/206) patients developed either immediate neurologic deficit (2.9%, 6/206) or delayed neurologic deficit (3.4%, 7/206). Of those who developed immediate neurologic deficit, none recovered. Of those who developed delayed neurologic deficit, 4/7 (57%) recovered neurologic function. Thus, 9 (4.3%) of 206 developed permanent neurologic deficit. 512 | www.annalsofsurgery.com
Late Survival For patients who completed both ET1 and ET2 procedures, the 1-, 5-, 10-, 15-, and 20-year survivals were 82.9%, 67.7%, 41.5%, 27.7%, and 13.6%, respectively. For patients who did not undergo completion repair, the 1, 5, 10, 15, and 20-year survivals were 56.1%, 39.3%, 31.2%, 21.1%, and 21.1%, respectively. Those who did not complete repair had a significantly reduced long-term survival compared with those who completed repair, P < 0.0001. (Fig. 2) The only independent risk factor for late death was age (hazard ratio = 1.025, P > 0.0001). The only independent factor that was protective against death in patients undergoing repair was completion repair, that is, those who underwent both ET1 and ET2 (hazard ratio = 1.832, P > 0.0001).
Incomplete Repair (Nonreturners)
The patients who did not return for the ET2 procedure (n = 131) were categorized into those who refused surgery (but were intended to undergo a second-stage repair) (32, 24.4%, median followup, 57.3 months), those who were unfit (30, 29.9%, median follow-up, 2.3 months), and those in whom the ET1 was performed prophylactically (69, 57.2%, median follow-up 2 months). Those who were either unfit for repair or refused surgery demonstrated a significantly worse survival prognosis when compared with those in whom ET was performed prophylactically (P > 0.0001) (Fig. 3). Of the 288 patients surviving the initial stage, 6 (2.2%) of 288 died within the 31- to 45-day optimal repair interval, suggesting that 6-week wait is safe. For the patients who were classified as unfit, the interval mortality was 17.8%.
Aortic Size Comparison of aortic sizes of the treated segment among nonreturners showed no statistical difference between the different subgroups, that is, unfit versus refusal versus prophylactic (P = 0.709). However, the aortic size of the unrepaired segment was significantly C 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
Outcomes of Elephant Trunk Technique
TABLE 1. 30-Day Mortality Risk Analysis: The ET1 Procedure Patient Variable
No. Patients, N (%)
Dead,∗ N (%)
297 (100)
25 (8.4)
75 (25.3) 74 (24.9) 74 (24.9) 74 (24.9) 140 (47.1) 157 (52.9) 17 (5.7) 61 (20.5)
5 (6.7) 5 (6.8) 6 (8.1) 9 (12.2) 14 (10.0) 11 (7.0) 3 (17.7) 8 (13.1)
75 (25.2) 74 (24.9) 74 (24.9) 74 (24.9) 28 (9.4) 119 (40.1) 73 (24.5) 80 (26.9) 245 (82.5) 73 (24.6) 281 (94.6) 138 (46.3)
8 (10.7) 7 (9.5) 6 (8.1) 4 (5.4) 4 (14.3) 9 (7.6) 6 (8.2) 7 (8.8) 20 (8.2) 8 (10.9) 23 (8.2) 10 (7.3)
75 (25.2) 74 (24.9) 76 (25.6) 72 (24.2)
7 (9.3) 5 (6.8) 9 (11.8) 4 (5.6)
75 (25.2) 76 (25.6) 73 (24.6) 73 (24.6)
7 (9.3) 4 (5.3) 6 (8.2) 8 (10.9)
Total Age,∗ yrs ≤58.47 58.47–67.01 67.01–73.36 >73.36 Females Males Emergent Preoperative renal failure Preoperative GFR∗ ≤54.66 54.66–71.82 71.82–92.25 >92.25 Acute dissection Chronic dissection Coronary Artery Disease COPD Systemic Hypertension Redo-median sternotomy Retrograde cerebral perfusion Aortic valve procedures Pump time, min∗ ≤123 123–149 149–179 >179 Clamp time,∗ min ≤65 65–81 81–104 >104
OR
CI
P
—
—
0.59
1.47
0.65–3.36
0.41
2.51 1.94
0.67–9.41 0.79–4.75
0.16 0.19
—
—
0.68
1.97 0.83 0.97 1.06 0.83 1.49 0.62 0.76
0.62–6.21 0.35–1.94 0.37–2.52 0.43–2.64 0.29–2.34 0.62–3.63 0.13–2.92 0.33–1.75
0.27 0.83 1.00 1.00 0.78 0.34 0.63 0.54
—
—
0.52
—
—
0.64
∗
Variables analyzed as quartiles. CI indicates confidence interval; COPD, chronic obstructive pulmonary disease.
smaller in the prophylactic group than in the unfit or refusal group (P < 0.0001) (Fig. 4).
DISCUSSION Although modifications to the original ET procedure have been made since its conception by Borst et al in 1983, the concept of placing a conduit into the descending thoracic aorta for later use or access was innovative and continues to be an important tool in the armamentarium of the aortic surgeon. Many of current challenges of EAA relate to multiorgan damage and in specific neurologic injury such as stroke or paraplegia. Advancements have been made regarding neurologic protection as was demonstrated in this report with a stroke incidence of 3.5% from ET1 procedure and an incidence of immediate neurologic deficit of 2.9% after ET2 procedure. Regarding ET1, despite growing preference for antegrade cerebral perfusion during circulatory arrest, we continue to espouse monitored RCP via the superior vena cava based on these continued noteworthy results.21 The principle adjunct for our approach to ET2 remains the use of cerebrospinal fluid drainage, distal aortic perfusion, and moderate hypothermia. Results from this report remain consistent with our previous reports with thoracoabdominal aortic aneurysm repair.20 Challenges still remain with postoperative renal dysfunction and the ET2 repair appears associated with higher rates of complications when compared with the ET1 repair.13,22 Hence, this supports our bias to perform the ET1 initially followed by the ET2 in the asymptomatic patient because recovery from the ET1 seems to be less complicated. Whether this is related to procedure-specific stress upon the patient or the relatively C 2014 Lippincott Williams & Wilkins
short time period between extensive procedures remains unknown. Considering this, the analysis of the time period between procedures became essential because a prolonged interval period may be associated with significant mortality as was noted in this series.16 One of the difficulties in analyzing this data set related to the definitions of the specific procedures, for example, ET1 and ET2, and the sequence of when the procedure was performed. In our previous series, we only analyzed a specific subset of ET patients, that is, the patient that we performed the ET1 and then the ET2 in sequence.16 Since that report, other variations of the ET procedure have been performed including the frozen ET, reversed ET, and hybrid ET that has led to differing sequences of the ET procedure. This report detailed all patients in our series that involved a staged ET procedure that was completed or was intended to be completed. For these reasons, the procedures were designated as ET1 and ET2, on the basis of approach and exposure, and the sequence was defined as either initial or completion. These designations aided in clarifying the less frequent occurrence of a patient undergoing the ET2 initially with an rET and later completing the ET1 with utilization of the reversed ET. At any rate, the majority of cases underwent the ET1 initially followed by the ET2 for completion. This remains our preferred sequence unless, the distal segment, that is, descending thoracic aorta, is much larger in size or associated with symptoms or rupture. Expanding on the analysis of our previous work,16 this study examined the subgroup of patients that did not complete the second repair, defined as the nonreturners. It was evident and intuitive that patients in whom the ET1 was performed prophylactically fared www.annalsofsurgery.com | 513
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Annals of Surgery r Volume 260, Number 3, September 2014
Estrera et al
TABLE 2. 30-Day Mortality Risk Analysis: The ET2 Procedure Patient Variable Total Age,∗ yrs ≤54.13 54.13–64.08 64.08–70.22 >70.22 Females Males Emergent Preoperative renal failure Preoperative GFR∗ ≤54.94 54.94–75.59 75.59–105.72 >105.72 Chronic dissection Coronary artery disease COPD Systemic Hypertension Diabetes Mellitus Preoperative stroke Rupture Redo thoracic aneurysm TAAA extent Extent 1 Extent 2 DTAA ∗ Pump time,∗ min ≤16 16–37 37–72 >72 Clamp time,∗ min ≤26 26–39 39–66 >66
No. Patients, N (%)
Mortality,∗ N (%)
206 (100)
21 (10.2)
52 (25.2) 51 (24.8) 52 (25.2) 51 (24.8) 95 (46.1) 111 (53.9) 10 (4.9) 38 (18.5)
4 (7.7) 4 (7.8) 4 (7.7) 9 (17.7) 8 (8.4) 13 (11.7) 2 (20.0) 6 (15.8)
52 (25.2) 51 (24.8) 52 (25.2) 51 (24.8) 108 (52.4) 50 (24.3) 58 (28.2) 169 (82.0) 19 (9.2) 15 (7.3) 12 (5.8) 9 (4.4)
8 (15.4) 6 (11.8) 4 (7.7) 3 (5.9) 12 (11.1) 6 (12.0) 6 (10.3) 19 (11.2) 4 (21.1) 1 (6.7) 5 (41.7) 2 (22.2)
55 (26.7) 48 (23.3) 103 (50.0)
6 (10.9) 4 (8.3) 11 (10.7)
57 (27.7) 48 (23.3) 51 (24.8) 50 (24.3)
9 (15.8) 2 (4.2) 3 (5.9) 7 (14.0)
58 (28.2) 48 (23.3) 51 (24.8) 49 (23.8)
4 (6.9) 3 (6.3) 7 (13.7) 7 (14.3)
OR
CI
P
—
—
0.25
0.69
0.27–1.75
0.49
2.33 1.91
0.46–11.77 0.69–5.31
0.27 0.23
—
—
0.38
1.24 1.28 1.02 2.22 2.67 0.61 7.95 2.68
0.49–3.07 0.47–3.50 0.38–2.78 0.49–9.96 0.79–8.95 0.08–4.89 2.26–27.92 0.51–13.81
0.82 0.59 1.00 0.38 0.11 1.00 0.003 0.23
—
—
0.89
—
—
0.13
—
—
0.38
∗
Variables analyzed as quartiles. CI indicates confidence interval; COPD, chronic obstructive pulmonary disease; DTAA, descending thoracic aortic aneurysm; TAAA, thoracoabdominal aortic aneurysm extent.
TABLE 3. Perioperative Morbidity Comparing ET1 Procedure With ET2 Procedure Patient Variable Total procedures (n = 491) Respiratory complications Postoperative dialysis Postoperative renal failure Myocardial Infarction GI complications Postoperative stroke Immediate neurological deficit Delayed neurological deficit
Ascending ± Arch, N (%)
Descending, N (%)
OR
CI
P
297 (59.1) 92 (30.9) 14 (4.7) 14 (4.7) 2 (0.7) 43 (14.5) 10 (3.4) 2 (0.7) 2 (0.7)
206 (40.9) 44 (21.4) 36 (17.5) 45 (21.8) 2 (0.9) 49 (23.8) 12 (5.8) 6 (2.9) 7 (3.4)
1.65 0.23 0.18 0.69 0.54 0.56 0.23 0.17
1.09–2.50 0.12–0.45 0.09–0.33 0.09–4.95 0.34–0.86 0.24–1.33 0.06–1.13 0.04–0.79
0.01
