Survival Benefit of Endovascular Descending Thoracic Aortic Repair for the High-Risk Patient Himanshu J. Patel, MD, Michael S. Shillingford, MD, David M. Williams, MD, Gilbert R. Upchurch, Jr, MD, Narasimham L. Dasika, MD, Richard L. Prager, MD, and G. Michael Deeb, MD Departments of Surgery and Radiology, University of Michigan Hospitals, Ann Arbor, Michigan Background. Despite acceptable results reported with endovascular thoracic aortic repair (TEVAR), recent studies have questioned the merit of repair in asymptomatic patients considered high risk for open surgery. In this group, advanced age or comorbid conditions may reduce life expectancy, thus limiting the benefit of elective aneurysmectomy. This study was conducted to determine whether elective TEVAR improves survival for this cohort. Methods. Forty-six asymptomatic patients with descending thoracic aortic disease were considered high risk for open surgery for reasons of age of 80 years or older (47.8%) or comorbid conditions (84.8%), and were subsequently evaluated for elective TEVAR. Of these, 21 underwent TEVAR, while another 25 patients were excluded from TEVAR on the basis of unfavorable anatomy or refused intervention. Results. The mean age of the cohort was 77.0 7.0 years (p 0.9 between groups). Prevalent comorbid conditions were similar between groups, and included coronary artery disease (p 1.0), chronic obstructive pulmonary disease (p 1.0), and peripheral vascular disease (p 0.23). Mean maximum aortic diameter was 6.0 1.4 cm (p 0.54 between groups). Indications for intervention included fusiform aneurysm (65.2%) and pseudoaneurysm or penetrating ulcer (32.6%). No 30-day mortality was observed after TEVAR. All-cause mortality in the entire cohort was 50%. Median actual time to mortality was different between groups (control, 9.2 months versus TEVAR, 24.9 months; p 0.01). Life-table analysis demonstrated improved survival for TEVAR at 24 months (p 0.05). Conclusions. Although the overall prognosis for the asymptomatic patient with descending thoracic aortic disease at high risk for open surgery is poor, elective endovascular repair improves survival and should be considered a therapeutic option in this setting. (Ann Thorac Surg 2007;83:1628 34) 2007 by The Society of Thoracic Surgeons Accepted for publication Dec 29, 2006. Presented at the Fifty-third Annual Meeting of the Southern Thoracic Surgical Association, Tucson, AZ, Nov 8 11, 2006. Address correspondence to Dr Patel, Department of Surgery, Section of Cardiac Surgery, 2120 Taubman Center Box 0348, Ann Arbor, MI 48109-0348; e-mail: hjpatel@med.umich.edu. Recent studies have demonstrated the feasibility of endovascular thoracic aortic repair (TEVAR) in a population considered high risk for an open surgical approach [1 5]. However, an update of the seminal work by Dake and associates [2] from Stanford University questioned the traditional indications for intervention for elective repair of thoracic aortic disease. In that study, the dismal 5-year prognosis identified for patients considered inoperable (31%) when compared with those deemed appropriate surgical candidates (78%) led the authors to question the need to intervene for the asymptomatic high-risk open surgical patient. Similar results have been suggested in trials of endovascular abdominal aortic repair (EVAR) [6]. The recent EVAR-2 trial, a randomized comparison of medical therapy versus EVAR for abdominal aortic aneurysms in patients considered unfit for open surgical procedures, suggested that there was no early survival advantage for this highrisk cohort. However, when analyzing the freedom from aneurysm-related mortality, a crossover point was present at 2 years, suggesting that longer follow-up may potentially lead to an improved survival in the treated cohort. The present study was conducted to determine whether elective TEVAR improves survival in the asymptomatic high-risk surgical cohort. All patients in this study were initially evaluated for TEVAR, and either medical or interventional treatment offered based solely on anatomic suitability for TEVAR or patient preference to forgo intervention. A survival analysis and comparison between groups was then performed to ascertain the merit of endovascular repair in this setting. Material and Methods This study was approved by the Institutional Review Board of the University of Michigan Hospitals. Informed consent requirements were waived for this study. Drs Patel, Williams, Upchurch, Dasika, and Deeb disclose that they have a financial relationship with Medtronic, Inc and Gore, Inc; Dr Patel also with Cook. 2007 by The Society of Thoracic Surgeons 0003-4975/07/$32.00 Published by Elsevier Inc doi:10.1016/j.athoracsur.2006.12.070
Ann Thorac Surg PATEL ET AL 2007;83:1628 34 SURVIVAL BENEFIT OF HIGH-RISK TEVAR Forty-six asymptomatic patients were prospectively identified at high risk for elective open descending thoracic aortic resection for reasons of age of 80 years or older or significant comorbid conditions, and constituted the study population. All patients were evaluated at the University of Michigan Hospitals from 1997 to 2005 by a thoracic surgeon with specific expertise in thoracic aortic reconstruction. This cohort was subsequently evaluated for TEVAR using a collaborative multidisciplinary approach (ie, both thoracic and vascular surgeons and interventional radiologists). A combination of factors, including available device characteristics, anatomic features of the aortic disease (ie, proximity to critical branch vessels and quality of proximal and distal landing zones), and quality and size of access vessels were used to determine feasibility and suitability for TEVAR. Indications for intervention included an absolute descending thoracic aortic size of 6 cm, an aortic diameter growth rate of more than 1 cm per year, or the diagnosis of a saccular aneurysm. Twenty-one patients underwent successful TEVAR. The remaining subset of patients (n 25) were either (1) thought to be unsuitable for TEVAR on the basis of anatomy (ie, need to cover the left carotid artery or celiac axis to achieve satisfactory proximal or distal landing zones, n 17); (2) refused any further intervention (n 7); or (3) had an unsuccessful TEVAR (n 1) without device deployment for successful aneurysm exclusion, and constituted the control group. No patient was excluded from TEVAR based on the lack of suitable access vessel for endograft deployment. Although most patients underwent TEVAR by means of a femoral or iliac route, 1 patient required delivery of the endograft by means of a partial upper sternotomy. For those patients undergoing TEVAR, endograft sizing was performed using spiral computed tomography with or without three-dimensional reconstruction, intravascular ultrasound, and calibrated angiography. Operative procedures were performed either in the operating room with fluoroscopy or in an angiography suite with fixed imaging equipment. General anesthesia was used in all. Percutaneous access was used to obtain necessary angiograms; the access vessel for endograft delivery was isolated through an open exposure. Device positioning and deployment were guided by angiographic landmarks or intravascular ultrasound. Completion aortography was performed, and all intraoperative type I or type III endoleaks were treated when identified by either repeat balloon dilatation to profile or, when necessary, additional coverage of the treated or adjacent aortic segments. Postoperative management for prevention of spinal cord ischemia was conducted according to standardized protocols. All lumbar drains were placed just after induction of anesthesia. Although no preoperative attempts were made to identify critical intercostal or lumbar vessels, 8 patients thought to be at high risk for spinal cord ischemia by the operating surgeon underwent lumbar drainage. General indications for lumbar drainage included the need to cover long aortic segments (eg, more than 20 cm), extensive coverage in the distal half of the descending thoracic aorta, or a previous history of abdominal aortic aneurysmectomy. No significant complications identified were directly attributed to lumbar drain placement. All patients were managed with mild permissive hypertension postoperatively to keep spinal perfusion pressures at 80 mm Hg or greater (if a lumbar drain was placed) or a mean arterial pressure of 90 to 100 mm Hg (if no lumbar drain was present). Duration of lumbar drainage was generally 24 to 48 hours. At that time, if no neurologic sequelae had occurred and the patient was hemodynamically stable, the drain was capped for an additional 6 to 8 hours before removal. The primary outcome of this study was all-cause mortality. Survival times were calculated from the point of patient contact when an indication for intervention was identified. The treated group underwent TEVAR at a median of 2.9 months after determination of the need for intervention. No patients experienced a rupture while waiting for their operative date. Follow-up was 100% Table 1. Demographics and Univariate Analysis of the Study Cohort Variable Control (n 25) 1629 TEVAR (n 21) p Value Demographics Age (years) 77.1 7.8 77.3 4.8 0.90 Number of patients 80 11 (44%) 7 (33%) 0.55 years of age Male sex 10 (40%) 8 (37%) 0.54 Maximum aortic 6.2 1.0 5.9 1.7 0.54 dimension (cm) Comorbidities CAD 15 (60%) 12 (57%) 1.0 History of CABG 6 (24%) 6 (28%) 0.75 COPD 12 (48%) 10 (46%) 1.0 Diabetes 4 (16%) 2 (9%) 0.67 Hypertension 25 (100%) 14 (67%) 0.003 Preoperative creatinine 1.6 1.1 1.2 0.4 0.12 (mg/dl) Prior AAA repair 10 (40%) 6 (28%) 0.54 PVOD 7 (28%) 10 (48%) 0.23 History of tobacco abuse 16 (73%) 14 (67%) 0.75 Aortic segments needing coverage Arch 13 (54%) 9 (43%) 0.55 Proximal descending 19 (76%) 19 (82%) 0.26 Distal descending 13 (52%) 18 (96%) 0.03 Total descending 7 (28%) 14 (67%) 0.007 Visceral segment 7 (29%) 0 (0%) 0.01 Aortic disease Fusiform aneurysm 15 (60%) 15 (71%) 0.54 Pseudoaneurysm or 10 (40%) 5 (24%) 0.35 penetrating ulcer Traumatic aortic injury 0 (0%) 1 (5%) 0.46 AAA abdominal aortic aneurysmectomy; CABG coronary artery bypass grafting; CAD coronary artery disease; COPD chronic obstructive pulmonary disease; PVOD peripheral vascular occlusive disease; TEVAR endovascular thoracic aorta repair. CARDIOVASCULAR
1630 PATEL ET AL Ann Thorac Surg SURVIVAL BENEFIT OF HIGH-RISK TEVAR 2007;83:1628 34 Table 2. Comorbidities Precluding Open Aortic Resection Comorbidity Precluding Open Repair a complete as of September 2006. Data were collected from clinic visit notes, hospital charts, and imaging studies, and mortality was verified by interrogation of the National Death Index. Duration of follow-up was a median of 18.9 months (mean, 24.8 19.8 months). Data were analyzed using SAS V8.2 (SAS, Cary, NC). Dichotomous variables were evaluated using 2 analysis; continuous variables using one-way analysis of variance. Survival was analyzed by life-table methods. All results with a p of 0.05 or less were considered statistically significant. Results Control (n 25) TEVAR (n 21) Nonreconstructable CAD with EF 0.40 1 1 Congestive heart failure 4 1 COPD with oxygen requirement, poor 8 13 pulmonary function, or severely limiting functional status Cerebrovascular disease with stroke 3 1 Cirrhosis with portal hypertension 1 1 Significant renal dysfunction (creatinine 3 0 2.5 mg/dl, or dialysis dependent) Functional status 4 5 Concomitant malignant neoplasm 3 2 Severe thrombocytopenia (platelet count 50,000) 1 1 a More than one may be present in a patient. CAD coronary artery disease; COPD chronic obstructive pulmonary disease; EF ejection fraction; TEVAR endovascular thoracic aorta repair. The mean age of the study cohort was 77.0 7.0 years (35% male). Comorbid conditions were prevalent and are included in Table 1. Mean maximum aortic diameter was 6.0 1.4 cm. Univariate analysis of demographic and comorbid conditions between groups is listed in Table 1. Indications for intervention included fusiform aneurysm (n 30, 63.8%), pseudoaneurysm or penetrating ulcer (n 16, 34%), and chronic traumatic aortic injury (n 1, 2.2%). The aortic segments that either needed coverage in untreated patients or those segments covered in treated patients are listed in Table 1. There were 22 patients (48%) considered high risk for open repair on the basis of age of 80 years or older. There was no difference in this frequency between groups (untreated, n 13 versus TEVAR, n 9; p 0.57). After evaluation by a thoracic surgeon with significant expertise in thoracic aortic reconstruction, 39 patients were designated not candidates for open surgery secondary to advanced comorbid conditions (eg, congestive heart failure, advanced chronic obstructive pulmonary disease). Although 20 patients (95.2%) in the TEVAR group met this criteria, 19 (76%) in the untreated group were similarly identified (p 0.11). The comorbid conditions precluding open repair are listed in Table 2. Endovascular Thoracic Aortic Repair Results For the treated patients, a median number of two endografts were used. The devices used included Gore TAG (W. L. Gore and Associates, Flagstaff, AZ) (n 11), Medtronic Talent (Medtronic, Inc, Minneapolis, MN) (n 8), and custom-fabricated devices in 2. Endovascular thoracic aortic repair was performed by means of a transfemoral route in 14 (66.7%), iliac exposure in 6 (28.6%), and from the ascending aorta through a partial upper sternotomy in 1. Concomitant procedures at the time of device implantation included ipsilateral iliac angioplasty to aid in device delivery in 2 and subsequent ileofemoral bypass in 2. One patient sustained an iliac artery rupture during attempted sheath removal and required insertion of a covered iliac stent graft for control of the hemorrhage. No in-hospital or 30-day mortality was observed after TEVAR. The median length of hospitalization was 8 days, with a median of 2 days spent in intensive care. No patients required dialysis, although 1 had transient renal failure (defined as a rise in creatinine greater than 2 Fig 1. Survival distribution function of the entire study cohort. This figure displays the survival curve by lifetable analysis for the entire cohort. In this cohort of highrisk inoperable patients, the overall all-cause mortality rate in the group is 50%, and life-table analysis demonstrates a 3-year survival of 56%.
Ann Thorac Surg PATEL ET AL 2007;83:1628 34 SURVIVAL BENEFIT OF HIGH-RISK TEVAR mg/dl). There were 2 patients who sustained temporary spinal cord ischemia, which responded to permissive hypertension and lumbar drainage. Only 1 patient had permanent spinal cord ischemia, on postoperative day 7, after presentation with delayed cardiac tamponade. This patient had endovascular repair performed through the ascending aorta by means of a partial sternotomy. Finally, 3 patients sustained a postoperative stroke. One patient experienced mild expressive aphasia on the second postoperative day and had near complete resolution of the defect by 4 months. Two patients had neurologic deficits on emergence from anesthesia one with expressive aphasia, and another with a right hemiparesis. Both patients likely had embolic events during TEVAR, and had near complete resolution of their defects by 1 and 4 months, respectively. Interestingly none of these patients sustaining postoperative stroke had coverage of the distal arch or the orifice of the left subclavian artery. In fact, 5 other patients had either partial or complete coverage of the left subclavian artery (with 2 of these undergoing prior planned left carotid to left subclavian artery bypass), and none of these 5 patients sustained a postoperative stroke. Therefore, no correlation was seen between extent of arch coverage and the occurrence of postoperative stroke. New or persistent endoleaks were observed in 3 patients (14.3%). The first had an endograft inserted in the distal arch at the region of high curvature, developed a proximal landing zone endoleak, and underwent another attempt at exclusion only to develop recurrent proximal type I endoleak 8 months later. Another patient who Fig 2. A comparison of survival for treated and untreated subgroups. This figure demonstrates a life-table analysis of the two subgroups in this study. It demonstrates that endovascular thoracic aorta repair (TEVAR) results in an early-term to intermediate-term survival advantage for the high-risk asymptomatic patient with elective thoracic aortic disease. Note that the 1- and 2-year life-table survival for endovascular thoracic aorta repair is 95% and 70%, respectively; in contrast, the 1- and 2-year life-table survival for the control arm is 68% (p 0.03) and 51% (p 0.05), respectively. Table 3. Causes of Mortality Cause of Mortality Untreated Control Group (n 25) TEVAR Group (n 21) Sudden death or suspected 4 3 aneurysm rupture Malignancy 0 1 Myocardial infarction 0 1 Cerebrovascular accident 0 1 Unknown cause 8 4 TEVAR endovascular thoracic aorta repair. 1631 required seven overlapping endografts to exclude the entire descending thoracic aorta developed a type III (junctional) endoleak 2 years later and could not undergo reintervention secondary to lack of a suitable access vessel. Because further therapy would have necessitated an open approach in these patients previously designated not suitable for open surgery, no additional treatment was offered. Both of these patients demonstrated sac growth and expired suddenly, likely as a result of aneurysm rupture. The third patient had a stable sac size, but a persistent type II endoleak that has decreased during the 12-month interval. Survival Analysis A survival curve generated by life-table analysis for the entire cohort is shown in Figure 1. The median survival time for the cohort by life-table analysis was 39.1 months. All-cause mortality in the entire cohort was 50% (n 23 patients), and did not differ between groups (untreated, n 13, 52% versus TEVAR, n 10, 48%; p 1.0). There were no preoperative variables (including age, sex, comorbid conditions, type of disease, or extent of repair needed) identified on univariate analysis that correlated with all-cause mortality (all p 0.1). The actual time to mortality was different between groups (untreated group, median time to mortality 9.2 months versus TEVAR, median time to mortality 24.9 months; p 0.01 by Mann- Whitney test). Finally, a comparison of survival curves for both the untreated and TEVAR groups is demonstrated in Figure 2. It is evident that there is a significant early force of mortality in the untreated group, but these curves are relatively parallel after the initial year. Although the overall mortality rate in both groups is similar, the survival curves show an overall early-term to intermediate-term survival benefit for TEVAR (p 0.05 at 24 months), likely as a result of a delay in time to mortality. Further information regarding the cause of death in the 22 expired patients is listed in Table 3. Autopsy results were not available for the majority of patients, and, therefore, a diagnosis of sudden death was identified as consistent with suspected aneurysm rupture. Given the significant early force of mortality, an assessment of cause of mortality in the 11 patients expiring in the first year was performed. Two of the 11 patients had undergone TEVAR and died of unknown causes. It is unlikely CARDIOVASCULAR
1632 PATEL ET AL Ann Thorac Surg SURVIVAL BENEFIT OF HIGH-RISK TEVAR 2007;83:1628 34 aneurysm rupture was the cause in both of them given their recent ( 4 months) antemortem imaging studies demonstrating successful aneurysm exclusion and stable sac dimension. The remaining nine patients were within the control group, and expired either of sudden death or aneurysm rupture (n 4) or unknown causes (n 5). This suggests that the significant differences in earlyterm and intermediate-term mortality between the control and treatment arms can be explained by aneurysm rupture. To clarify the causes of mortality during the entire study period in the untreated group, an assessment of aneurysm size was performed in control patients who expired versus those who survived to determine whether mortality was secondary to an overall larger aneurysm dimension. This did not reveal a significant difference (expired controls, mean aneurysm size 6.3 1.3 cm versus alive controls, 6.1 0.6 cm; p 0.55). Finally, of the 7 control patients who elected to forgo intervention and were not anatomic candidates for TEVAR, 1 expired of sudden death or aneurysm rupture and 2 died of unknown reasons. Comment The decision to proceed with elective aortic repair in the asymptomatic patient is predicated on the assumption that the benefit of intervention outweighs the risk of continued observation. Natural history studies on thoracic aortic aneurysmal disease have identified risk factors for rupture including maximum aortic diameter, growth rate, and age, as well as the presence of symptoms, hypertension, or chronic obstructive pulmonary disease [7 9]. In addition, prior studies have also determined that descending thoracic aneurysms have a poorer prognosis than those in the ascending aorta, likely because of the associated comorbidities present in this population [9]. Estimated risks of mortality, rupture, or dissection for patients with thoracic aneurysms greater than 6 cm have been reported in excess of 15% per year [9]. Endovascular approaches to the descending thoracic aorta have been successfully applied in patients considered high risk for open surgical repair [1 6]. However, recent studies have questioned the merit of elective repair in this cohort, suggesting that the advanced age or comorbid conditions elevating the risk of intervention may obviate the benefit afforded by aneurysm repair [2]. The recent report by Demers and colleagues [2] of TEVAR showed that the 5-year survival in patients suitable for open surgery was 78%, and this contrasted significantly with the poor 5-year survival (31%) in those deemed inoperable. Similar results have been obtained from studies of high-risk patients after EVAR. The observational data obtained from the EUROSTAR registry documented an actual 1-year mortality rate of greater than 20% in patients considered unfit for open surgery undergoing EVAR [10]. In contrast, the survival rates for those patients who were considered suitable for open repair had a significant survival advantage compared with the highrisk cohort. Based on these data, the authors suggested that EVAR may result in a limited survival benefit in the patient population with a significant comorbid burden. The prospective randomized EVAR Trial 2 compared medical therapy versus EVAR in patients at high risk for open surgery [6]. In that study, there was no significant survival benefit for endovascular repair at 4 years. An overall 9% perioperative mortality was reported with EVAR, with an overall survival for the treated group at 4 years of 34%. The medical arm reported a 4-year survival of 38%. However, a major criticism of EVAR-2 was that 52% of the perioperative mortality occurred in the time interval between randomization and the scheduled operative date. Other criticisms of that study included both a 22% crossover of the medical arm undergoing either subsequent EVAR or even open surgery. Finally, analysis of freedom from aneurysm-related mortality demonstrated a crossover point at 2 years. This suggested that longer-term follow-up may lead to an improvement in survival for the treated arm. To address the issues of excessive perioperative mortality and high crossover to treatment, a recent reevaluation of the Society of Vascular Surgeons database of the five Food and Drug Association approved multicenter trials for new abdominal endografts was performed [11]. That study compared survival in high-risk patients undergoing EVAR versus open surgery and found comparable rates of early, aneurysm-related, and late death in both arms, suggesting that EVAR was safe and durable in high-risk patients and that treatment options in this cohort should be individualized. In this study, all patients were prospectively identified as unsuitable for open repair because of advanced age or comorbid conditions (Table 2). This entire group was then considered for TEVAR on an intent-to-treat basis. All of these patients then underwent TEVAR unless excluded by anatomic criteria or patient preference to forgo intervention. Age, comorbidities, or other reasons did not play a role in eliminating patients from eligibility for endovascular therapy. Therefore, selection bias was not likely a significant factor accounting for differences in mortality between the treatment and observation arms in this study. In fact, both patient groups were similar with respect to age, mean aneurysm dimension, and aortic disease, as well as prevalence of associated comorbidities (with the exception of hypertension). In addition, more patients prospectively met high-risk criteria on the basis of comorbid conditions in the TEVAR arm (95%) than the control arm (76%; p 0.1), suggesting that any potential difference in survival was not likely caused by the presence of a more physiologically unfit control group. Despite the relatively similar profile between the groups, survival analysis (Fig 2) demonstrated a statistically significant reduction in all-cause mortality at 1 and 2 years. When an analysis of all expired patients was conducted, although the overall mortality rates were not different between groups, the median time to mortality was significantly less in the untreated group. Our results suggest that although the high-risk surgical cohort has a significant force of mortality (Fig 1), TEVAR delays the time to
Ann Thorac Surg PATEL ET AL 2007;83:1628 34 SURVIVAL BENEFIT OF HIGH-RISK TEVAR death, and may be viewed as a palliative therapy in this setting. It is important to note that in this study, the indications for intervention conform to accepted standard indications for operation in the asymptomatic patient with elective thoracic aortic disease. Patients were only evaluated for TEVAR if aneurysm dimension met absolute size criteria ( 6 cm), growth criteria ( 1 cm/year), or anatomic criteria (saccular aneurysm) for open surgery. This is particularly important as prior work from our group, as well as others, has demonstrated that TEVAR in high-risk cohorts does carry a risk of morbidity including death (2% to 5%), stroke (2% to 8%), dialysis (1% to 5%), and paraplegia (1% to 5%) [1 5]. Coupling the reported risk of rupture for smaller aneurysms (ie, 5 to 6 cm) [7 9] with the potential decrease in perioperative morbidity and mortality reported with TEVAR [1 5], an important future study could evaluate the survival advantage in treating smaller aneurysms in the high-risk patient. Limitations of this study are its small sample size, nonrandomized nature, and retrospective design. However, this cohort was initially prospectively identified at high risk for open surgery, and only separated into medical and interventional arms on the basis of either unfavorable anatomy or patient preference. Despite these limitations, there was a clear survival benefit established for the treated group. Another limitation of this study are the lack of available data of true aneurysmrelated mortality. However, all-cause mortality is likely a more suitable end point to evaluate TEVAR in this patient population, given the hypothesis that the associated comorbid conditions and not aneurysm-related mortality would likely limit patient survival in this group. In conclusion, this study demonstrates an early-term to intermediate-term survival advantage of elective TEVAR in the asymptomatic patient at high risk for open surgery. These results should encourage a prospective randomized trial comparing medical versus endoluminal strategies for the high-risk patient with elective thoracic aortic disease. We gratefully acknowledge the assistance of Mary Proctor, MS, for her help in the statistical analysis of the data. References 1633 1. Leurs LJ, Bell R, Degrieck Y, et al. Endovascular treatment of thoracic aortic diseases: combined experience from the EUROSTAR and United Kingdom Thoracic Endograft registries. J Vasc Surg 2004;40:670 80. 2. Demers P, Miller DC, Mitchell RS, et al. Midterm results of endovascular repair of descending thoracic aortic aneurysms with first-generation stent grafts. J Thorac Cardiovasc Surg 2004;127:664 73. 3. Greenberg RK, O Neill S, Walker E, et al. Endovascular repair of thoracic aortic lesions with the Zenith TX1 and TX2 thoracic endografts: intermediate-term results. J Vasc Surg 2005;41:589 96. 4. Katzen BT, Dake MD, MacLean AA, et al. Endovascular repair of abdominal and thoracic aortic aneurysms. Circulation 2005; 112:1663 75. 5. Patel HJ, Williams DM, Upchurch GR Jr, et al. Long-term results from a 12-year experience with endovascular therapy for thoracic aortic disease. Ann Thorac Surg 2006;82:2147 53. 6. EVAR Trial Participants. Endovascular aneurysm repair and outcome in patients unfit for open repair of abdominal aortic aneurysm (EVAR trial 2): randomized controlled trial. Lancet 2005;365:2187 92. 7. Griepp RB, Ergin MA, Galla JD, et al. Natural history of descending thoracic and thoracoabdominal aneurysms. Ann Thorac Surg 1999;67:1927 30. 8. Coady MA, Rizzo JA, Hammond GL, et al. What is the appropriate size criterion for resection of thoracic aortic aneurysms. J Thorac Cardiovasc Surg 1997;113:476 91. 9. Davies RR, Goldstein LJ, Coady MA, et al. Yearly rupture or dissection rates for thoracic aortic aneurysms: simple prediction based on size. Ann Thorac Surg 2002;73:17 28. 10. Laheij RJF, van Marrewijk CJ. Endovascular stenting of abdominal aortic aneurysm in patients unfit for elective open surgery. Eurostar group. EUROpean collaborators registry on Stent-graft Techniques for abdominal aortic Aneurysm Repair. Lancet 2000;356:832. 11. Sicard GA, Zwolak RM, Sidawy AN, et al. Endovascular abdominal aortic aneurysm repair: long term outcome measures in patients at high risk for open surgery. J Vasc Surg 2006;44:229 36. CARDIOVASCULAR DISCUSSION DR JOHN A. KERN (Charlottesville, VA): Thank you, Doctor Patel. That was a very nice paper and you should be congratulated on your results in this incredibly difficult group of patients to manage. Your results of endovascular therapy in this highrisk patient population really parallels our own. What you found is it can be done; it can be done safely. There are some risks associated. Even though the risk of things like paraplegia and stroke were small, they are still real and I think it needs to be taken into consideration. The biggest limitation of this study is it clearly is not a prospective randomized trial. You mention that you really didn t affect overall mortality but the time to death was longer. The problem is in the patients not undergoing endovascular therapy, the reason that they didn t undergo that therapy was either they declined or they weren t anatomic candidates for the reasons you pointed out, but that could clearly translate into the fact that their aortic pathology was such that it tended to put them in a group that was going to die sooner. What I am getting at is I am not sure that endovascular therapy is the fountain of youth for these folks, and I guess what I am trying to get at is, based on these data, how do you now decide who to proceed with therapy given the fact that they are anatomic candidates? The second question, the one thing we found in our group of patients is there tended to be for us an association of female gender to an increased rate of postoperative complications such as prolonged ventilatory support, renal dysfunction, and we don t know why that is; presumably maybe longer operative times required for gaining access and things like that. Did you notice that similar finding in your group of patients? Thank you. DR PATEL: Thank you, Doctor Kern, for your questions. To answer your second question first, we did not identify any effect of gender on postoperative outcome in this study. In a previous
1634 PATEL ET AL Ann Thorac Surg SURVIVAL BENEFIT OF HIGH-RISK TEVAR 2007;83:1628 34 report of our total endovascular experience recently presented at the STS (Society of Thoracic Surgeons), we also did not identify sex as a risk factor for postoperative major morbidity or mortality. In response to your other questions regarding the nature of this study, it is true that this was a retrospective analysis. However, what should be noted is that these patients were prospectively identified as high risk for open repair, and on that basis, all of them were evaluated for TEVAR (endovascular thoracic aorta repair) on an intent-to-treat basis. The formulated control arm consisted only of those patients who either refused an intervention after hearing the risks or benefits, or those patients who were not considered TEVAR candidates because of need for coverage of critical branch vessels. Currently, when we evaluate patients who we feel are high risk for open operations, we present them with all the options. We tell them that they may expire from other comorbid conditions, but TEVAR may ameliorate mortality from aneurysmrelated death and thus give them a short-term to intermediateterm survival advantage. Unfortunately, we do not have a defined aneurysm-related mortality in the current report, which I think would be a very important aspect for a future study. However, I think what this study is designed to do is to explore the stated hypothesis and perhaps suggest that the time is now very ripe for a prospective randomized multicenter trial to evaluate endovascular therapy for these patients considered at high risk for open surgery, similar to what the EVAR-2 trial was. DR THORALF SUNDT (Rochester, MN): Well, first I would like to congratulate you on starting your presentation with a quote from William Osler. So I take my hat off to you for that. The issue of high risk is in the eye of the beholder, and we see this all the time with the cardiologists who decide that they are going to do a percutaneous intervention instead of a surgical intervention because somebody is high risk. That will be an issue to be resolved with a randomized study of no intervention versus stent grafting as well. So I am a little bit concerned about how you defined who was high risk, especially when you talk about an operation, as we heard yesterday from Joe Coselli, that is high risk. These operations are high risk in anybody. Specifically, did you exclude everybody over the age of 80? Is just the age of 80 considered a contraindication to open repair in your institution? DR PATEL: No, the fact that you are over 80 years old alone does not contraindicate open surgery on its own at our institution. These patients were first evaluated by a thoracic surgeon with expertise in aortic reconstruction, and specifically felt to be high risk for open surgery. What was evident was about 48% were considered high risk for advanced age, but 85% of these patients were considered high risk for secondary comorbid conditions. DR SUNDT: What was the cost associated with this intervention? DR PATEL: Unfortunately we don t have cost data.