Patients undergoing coronary artery bypass grafting (CABG) with ascending aortic atheromatous disease are known to carry increased risk of death, stroke, and major adverse cardiac and cerebrovascular events (MACCE).1–3 Advanced age, female sex, previous stroke, transient ischemic attack, carotid bruits, hypertension, diabetes mellitus, congestive heart failure, and extensive aortic calcification have all been associated with increased risk of operative strokes.4–6 As our understanding of stroke predictive factors has increased, more attention is being focused on the observation that most strokes are embolic in nature and are discovered early after surgery (≤24 hours postoperative).4 Numerous diagnostic techniques have been proposed to identify patients who are at increased risk of aortic atherosclerotic changes. Modalities such as noncontrast computed tomography, transesophageal echocardiography (TEE), and intraoperative epiaortic ultrasound (EU) have all been shown to detect the atheromatous aorta with differing abilities.7–10
The intraoperative use of EU is not novel or new. In fact, it has been used in the cardiac operating rooms for years. Initially, it was used to evaluate valvular pathology intraoperatively,11 but has since been replaced by TEE. In recent years, EU has been shown to be superior to TEE in the evaluation of the atheromatous ascending aorta,8,10 as the distal ascending and transverse arch views in TEE are obstructed by the mainstem bronchi.12,13 Few studies utilizing EU for aortic screening have suggested improvements in outcome as a result of intraoperative modifications of the surgical plan.14,15 Most of these studies are small, thus confounding data interpretation. However, a recent study by Rosenberger et al16, which evaluated over 6000 patients with EU, suggested that the operative course was changed in 4% of patients due to the finding of aortic pathology, resulting in improved neurologic outcomes.
To reduce the risk of operative-related strokes in this patient population, surgeons have started to use methods to reduce interference with plaque-laden aortas. The use of off pump coronary bypass, echo-guided cannula/clamp placement, and no touch aortic procedures, among others, have been the most widely used. In our practice, identification of the patient with an atheromatous aorta has become paramount. EU is part of the operative plan in all patients, and the conduct of the operation is altered based on its findings. As a result, we have investigated our recent experience to help further delineate the benefit of intraoperative EU in patients undergoing isolated, primary CABG.
PATIENTS AND METHODS
The institution’s Society of Thoracic Surgeons (STS) Adult Cardiac Database was searched for all patients who underwent primary CABG between January 1, 2004 and March 30, 2007. This time frame was chosen to coincide with the period during which epiaortic grade was collected routinely. The cohort consisted of 5198 consecutive patients of which 4278 (82.3%) had EU grade collected and constitutes the study population for the current study. Database records from this retrospective, single-center cohort study included demographic data, comorbidities, surgeon identity, operative strategy, and clinical outcomes. This study was approved by the Institutional Review Board and was conducted in compliance with Health Insurance Portability and Accountability Act (HIPAA) regulations and the Declaration of Helsinki. Individual consent was waived by the Institutional Review Board.
Off-pump coronary artery bypass (OPCAB) was performed with one of several commercially available cardiac positioning and coronary artery stabilizing devices, using techniques that have been previously described.17 Conventional CABG on cardiopulmonary bypass was used with standard techniques, utilizing roller head pumps, membrane oxygenators, cardiotomy suction, arterial filters, cold antegrade and retrograde blood cardioplegia, and moderate systemic hypothermia (30°C-34°C). Aortic occlusion, when used, was achieved using either a complete or a partial aortic cross clamp.
EU has been performed at our institution routinely since 2004 using a modification of Katz et al.18 A 12-MHz transthoracic echocardiography probe with an acoustic window is placed into a sterile saline filled bag and used for aortic scanning immediately after the pericardial well has been opened and filled with saline. Images are acquired in the short-axis view of the proximal aorta, the mid aorta at the level of routine proximal anastomosis, and the distal ascending aorta at the level of cross clamp placement. Further images are taken of the arch utilizing a long-axis view. If TEE was used in the case, a grade of descending aorta would also be given. Grade was determined based on thickness of the aortic wall. Grade 1 aortas had minimal intimal thickening, whereas grade 2 aortas had intimal thickening up to 3 mm. Grade 3 aortas had between 3 and 5mm of thickening, where grade 4 aortas had more than 5 mm of thickening. Grade 5 aortas possessed mobile atheroma. The highest grade of all views of the ascending aorta was then denoted to be the ascending aortic grade.
Before data analysis, preoperative risk factors for the outcomes of interest were identified and harvested from the institutional STS database (Table 1). These variables were chosen because of their potential for contributing to selection bias. Standard STS definitions for each risk factor and outcome were used (https://www.sts.org). Patients were classified primarily by two variables—epiaortic grade (grades 1–2 or 3–5, grade) and whether they were clamped (yes or no, Clamped). Patients that received a partial cross clamp were considered to have been clamped and grouped with patients that had the full aortic cross clamp. Additionally, the EU data extraction sheet has a field that indicates whether the findings of the ultrasound influenced the surgeon’s choice of surgical approach. Race was dichotomized to either white or non-white. The adverse outcomes examined were operative mortality, permanent stroke, myocardial infarction, and MACCE. The medical records STS database was populated by trained and exclusively dedicated personnel; thus, missing data were scarce. Data were 100% complete for the primary variables of interest—grade and Clamped—as well as for each major postoperative hospital outcome. Eight variables had missing data; height (n = 4, 0.1%), weight (n = 1, <0.1%), ejection fraction (n = 128, 3.0%), number of diseased vessels (n = 5, 0.1%), STS predicted risk of mortality (n = 463, 10.8%), last creatinine level (n = 16, 0.4%), Canadian Cardiovascular Society angina classification (n = 65, 1.5%), and New York Heart Association heart failure classification (n = 64, 1.5%).
Data quality checks were used both at the institutional level and before final entry into the STS adult cardiac database. A multiple imputation algorithm was used to impute values that reflect the uncertainty surrounding the missing data. This algorithm was not performed in an effort to recreate the truth; rather, the goal was to avoid selection bias that can occur by deleting cases with missing variables of interest. Five datasets were multiply imputed and parameter estimates from the five datasets were combined using methods described by Molenberghs and Kenward.19 Data were assumed to be missing at random.
Patients were classified according to grade and Clamped. To control for potential selection bias in which patients get clamped, propensity scoring methods, popularly described by Blackstone,20 were calculated for each patient from 45 risk factors available preoperatively including surgeon identity, year of surgery, and eight indicators of missingness (Table 1). To estimate the propensity score (PS), a multiple logistic regression model was used nonparsimoniously to model Clamped (Yes or No) as a function of all 45 risk factors (Table 1). The resulting conditional probability of a patient being clamped is the PS. The goal of the PS adjusts the groups with respect to their preoperative risk factors so that nonconfounded comparisons of group effects can be estimated in an unbiased fashion. The PS was then used as a regression covariate in the final logistic regression models (see below). The underlying assumption of a linear relationship between the PS and the logit of the predicted probabilities of each outcome was verified by plotting these quantities in deciles.
To statistically evaluate the effects of grade and Clamped, multivariable logistic regression models were constructed for each adverse outcome. Each model consisted of Clamped, grade, and their interaction adjusted for the PS. If the interaction was statistically insignificant then the interaction term was removed from the model, and the main effects alone were evaluated. A significant interaction would indicate that the effect of clamping was different in patients of grades 1 to 2 than it was in patients of grade 3 to 5. Adjusted odds ratios (AOR) associated with Clamped and grade, along with 95% confidence intervals, were calculated for each of the four adverse outcome endpoints. All logistic models were adjusted with the PS.
The data were managed and analyzed using SAS Version 9.1 (Cary, NC). Unadjusted comparisons were performed with χ2 tests and two-sample t-tests for categorical and continuous predictors, respectively. All statistical tests were two-sided using a P = 0.05 level of significance. No adjustments for multiple tests were made.
Demographic and risk stratification data are presented in Table 1. During the period of study, 4278 patients underwent primary, isolated CABG and EU screening of the ascending aorta. Eighty-six percent of patients had an aorta considered low risk for neurologic event by EU (n = 3,686, grade 1–2) and 14% had a high-risk aorta (n = 592, grade 3–5). The breakdown of epiaortic grade was as follows: grade 1: 2558 (59.7%), grade 2: 1135 (26.5%), grade 3: 455 (10.6%), grade 4: 85 (2%), and grade 5: 52 (1.2%). Three techniques of aortic occlusion were used: (1) two-clamp technique (full aortic cross clamp followed by partial side-biting clamp), (2) partial side-biting clamp, or (3) no aortic cross clamp. In those with no aortic cross clamp, a variety of techniques were used for construction of proximal anastomoses including Heartstring devices (Maquet, San Jose, CA) or sequential arterial grafting. In grade 1 to 2, 35.7% (n = 1314) of patients underwent a two-clamp technique, where 54.5% (n = 2011) of patients underwent a side-biting clamp and 9.8% (n = 361) of patients underwent no clamping. In grade 3 to 5, 26% (n = 154) of patients underwent a two clamp technique, where 19.4% (n = 115) of patients underwent a side-biting clamp, and 54.6% (n = 323) of patients underwent no clamp. As expected, patients in grade 3 to 5 tended to be older, male, had more cerebrovascular disease, suffered more preoperative strokes, and had a higher incidence of peripheral vascular disease. These patients were more often current smokers at the time of operation, had more significant chronic obstructive pulmonary disease and a higher incidence of preoperative renal failure. The STS estimate of predicted mortality was also higher in patients with a grade 3 to 5 aorta than those in grade 1 to 2 (5% vs. 2.8%, P < 0.001).
These data are the result of a retrospective review of patients undergoing cardiac surgery from January 2004 to March 2007. Unadjusted outcomes were compared by aortic grade and use of clamping (Table 2). The incidence of death was higher in grade 3 to 5 when a clamp was applied to the aorta as compared with a no clamp technique (7.1% vs. 2.8%: P = 0.015). The incidence of death was also increased in patients with a grade 3 to 5 aorta that was clamped compared with those with a grade 1 to 2 aorta that was clamped (7.1% vs. 2.1%: P < 0.001). Stroke was also increased in patients with grade 3 to 5 nonclamped aortas compared with patients with grade 1 to 2 (1.9% vs. 0.3%; P = 0.04) nonclamped aortas. Correspondingly, MACCE was elevated in patients with grade 3 to 5 clamped aortas compared with no clamp (9.7% vs. 5.3%; P = 0.04), as well as patients with a grade 3 to 5 clamped aorta compared with patients with grade 1 to 2 clamped aorta (9.7% vs. 3.7%, P < 0.001).
When the data was adjusted for propensity scoring (Table 3), patients with a grade 3 to 5 aorta undergoing CABG suffered an increased risk of death (AOR 3.11; P < 0.001), stroke (AOR 2.12; P < 0.001), and MACCE (AOR 2.58; P < 0.001). Furthermore, the placement of an aortic clamp (any clamp, all grades) led to almost a threefold higher risk of stroke (AOR 2.77; P = 0.032).
Of note, data was specifically collected as to whether the findings of EU caused the surgeon to change the operative plan for revascularization. In 12.4% (n = 530) of patients, the operating surgeon reported that the EU findings led to a change in operative strategy. Outcomes based on the changed operative strategy are noted in Table 4. In the group where surgeon strategy was changed, there were no differences observed in death for patients with grade 3 to 5 clamped aortas compared with no clamp (4.8% vs. 2.7%, P = 0.29) as well as patients with grade 3 to 5 clamped aortas compared with grade 1 to 2 clamped (4.8% vs. 2.6%, P = 0.45). Furthermore, no differences in death were noted between patients with grade 3 to 5 aortas and grade 1 to 2 aortas without the application of a clamp (2.7% vs. 2.0%, P = 0.78). When stroke rates are examined, no statistically significant differences were noted when patients with grade 3 to 5 aortas were compared with grade 1 to 2 aortas with (3.9% vs. 0%, P = 0.08) or without a cross clamp (2.0% vs. 2.0%, P = 0.30). There were no noticeable differences between groups with respect to myocardial infarction. Finally, with respect to MACCE, no differences were observed for patients with grade 3 to 5 clamped aortas compared with no clamp (7.7% vs. 5.4%, P = 0.38) as well as patients with grade 3 to 5 clamped aortas compared with grade 1 to 2 clamped (7.7% vs. 3.9%, P = 0.29).
In patients in which EU did not alter surgeon strategy, the incidence of death was increased in the grade 3 to 5 clamped group when compared with the grade 1 to 2 clamped group (8.5% vs. 2.1%, P < 0.001). The incidence of MACCE was also increased in the grade 3 to 5 clamped group compared with the grade 1 to 2 clamped group (10.9% vs. 3.7%, P < 0.001). In the remaining comparisons of groups where surgeon strategy was not altered, no other differences were noted.
Much discussion has occurred about methods to reduce the incidence of stroke after CABG. These methods include screening of high-risk individuals,5 avoidance of cardiopulmonary bypass, and the utilization of OPCAB techniques.21–24 Other methods to reduce stroke include echocardiographic guidance of aortic cannulation, cross clamp placement, and proximal anastomotic localization,25,26 as well as the total avoidance of aortic cross clamping in the atheromatous ascending aorta.14,15,27–29 Some have advocated the use of hypothermic fibrillatory arrest,14,16 as well as, ascending aortic replacement for the significantly diseased proximal aorta.30 Utilization of extra-anatomic sites or the liberal utilization of sequential proximal anastomoses, such as T or Y grafting from arterial conduits, has also been described to reduce aortic manipulations.26 Newer methods advocated to reduce the incidence of stroke include the use of proximal anastomotic devices that avoid the use of an aortic clamp and minimize aortic manipulation (eg, Heartstring [Maquet, San Jose, CA] or PAS-port [Cardica, Redwood City, CA]).
To date, the main focus of reducing the incidence of stroke associated with coronary revascularization has been identifying and avoiding placement of an aortic clamp on the atheromatous aorta. With published reports indicating that up to 75% of all cerebrovascular accidents after CABG are embolic and early in nature,4 it becomes critical to make every effort to reduce the risk of distal embolization from aortic sourced atheroma. Most cerebral emboli have been shown to occur at the time of cross clamp removal,31and side biting clamps have been shown to have the same stroke risk as full cross clamps.32 Methods of identifying the atheromatous aorta include noncontrast computed tomographic scanning of the chest and transesophageal echocardiographic identification of atheromatous deposits in the aortic wall. Intraoperative EU has been shown to be superior to TEE or palpation alone in identifying aortic atheromatous lesions, especially in the mid to distal ascending aorta,8–10 thus making it the modality of choice in identifying ascending aortic atheroma.
In our study, we found that EU identified 86% of patients as having an aorta considered low risk for neurologic sequelea (grade 1–2) and 14% with a high-risk aorta (grade 3–5). These findings led to a change of the surgical plan in 12% of cases. The alterations in plan included changing to an OPCAB approach, changing cannulation/clamp/proximal placement sites, conversion from a two clamp technique to a single clamp, as well as conversion to no clamp techniques. These data are consistent with other published reports of alteration in surgical strategy from smaller studies,14,33 but are significantly higher than that found by Rosenberger et al, who found their strategy was altered in only 4.1% of over 6000 patients undergoing all types of cardiac surgery, and 4.6% (162 of 3535) of those undergoing CABG. They also reported that in patients who underwent modifications in technique, there was a decreased incidence of stroke.16 Davila-Roman et al34 also reported that there was >1.5-fold increase in the incidence of both neurologic events and mortality as the severity of atherosclerosis increased from normal-mild to moderate, and a greater than threefold increase in the incidence of both as the severity of atherosclerosis increased from normal-mild to severe.
In this study, our results indicate that aortas with high EU grade undergoing CABG, had an increased risk of mortality, stroke, and MACCE as compared with patients with low-grade aortas. Of further interest, we found that in patients with a high-risk aorta, the placement of a clamp led to an increased risk of mortality nearly threefold higher than in those patients considered low risk. Perhaps, the most intriguing finding of this study is the apparent reduction in the incidence of mortality and MACCE in patients who underwent a change in operative strategy based on the EU findings. This is most evident in the groups that used an aortic clamp, where the incidence of death was similar in grade 3 to 5 clamped aortas compared with grade 1 to 2 clamped aortas (4.8% vs. 2.7%; P = 0.29). Additionally, in patients who underwent an operative strategy change, the incidence of MACCE was similar in grade 3 to 5 clamped aortas compared with grade 1 to 2 clamped aortas (7.7% vs. 5.4%, P = 0.38).
This study has several limitations. It is a retrospective and nonrandomized, review of a single institution’s practice. There was no comparable control group that did not receive EU. We feel that the preexisting data already discussed on the benefit of EU is sufficient to use it on all patients. This may introduce some bias, but we feel that the change in the operative plan only benefits the patient with a high-risk aorta. Furthermore, it should also be noted that all of our surgeons are proficient in the use of OPCAB and clampless proximal anastomoses, but preferences for one technique over another persist. Finally, for most of our surgeons, OPCAB is the procedure of choice for primary coronary revascularization.
As discussed by Rosenberger et al,16 complications of EU are rare. Because the probe is placed in a sterile, fluid filled bag, any perforation of the bag may lead to contamination of the surgical field. We have not found this to be a problem and think it of little risk. There is some argument about the length of time required to perform an examination, but we have found that our exams rarely take more than a few minutes. Some argue that the interpretation of the ultrasound is time consuming and difficult to learn but, the American Society of Echocardiography and Society of Cardiovascular Anesthesiologists both recommend this training as a core component of their advanced certification.35
In summary, we feel that EU should become the gold standard in the evaluation of the ascending aorta and aortic arch in cardiac surgery. Our data clearly indicates that patients with a high grade EU rating are at increased risk of death, stroke, and MACCE. Every effort should be made to avoid clamping a grade 3 to 5 aorta, but EU- guided intervention may limit morbidity and mortality if a clamp is a necessity.
1.Schachner T, Zimmer A, Nagele G, et al. The influence of ascending aortic atherosclerosis on the long term survival after CABG. Eur J Cardiothorac Surg
2.Hogue CW Jr, Murphy SF, Schechtman KB, et al. Risk factors for early or delayed stoke after cardiac surgery. Circulation
3.Das S, Dunning J. Can epiaortic ultrasound reduce the incidence of intraoperative stroke during cardiac surgery? Interact Cardiovasc Thorac Surg
4.Filsoufi F, Rahmanian PB, Castillo JG, et al. Incidence, topography, predictors and long term survival after stroke in patients undergoing coronary artery bypass grafting. Ann Thorac Surg
5.Puskas JD, Winston AD, Wright CE, et al. Stroke after coronary artery operation: incidence, correlates, outcome and cost. Ann Thorac Surg
6.Halkos ME, Puskas JD, Lattouf OM, et al. Impact of preoperative neurologic events on outcomes after coronary artery bypass grafting. Ann Thorac Surg
7.Lee R, Matsutani N, Polimenakos AC, et al. Preoperative noncontrast chest computed tomography identifies potential aortic emboli. Ann Thorac Surg
8.Sylivris S, Calafiore P, Matalanis G, et al. The intraoperative assessment of ascending aortic atheroma: epiaortic imaging is superior to both transesophageal echocardiography and direct palpation. J Cardiothorac Vasc Anesth
9.Bolotin G, Domany Y, de Perini L, et al. Use of intraoperative epiaortic ultrasonography to delineate aortic atheroma. Chest
10.Suvarna S, Smith A, Stygall J, et al. An intraoperative assessment of the ascending aorta: a comparison of digital palpation, transesophageal echocardiography, and epiaortic ultrasonography. J Cardiovasc Vasc Anesth
11.Edrich T, Shernan SK, Smith B, et al. Usefulness of intraoperative epiaortic echocardiography to resolve discrepancy between transthoracic and transesophageal measurements of aortic valve gradient. Can J Anesth
12.Phoon CK, Rutkowski M. Transesophageal imaging of the mid to distal left pulmonary artery in congenital heart disease. J Am Soc Echocardiogr
13.Rosenberger P, Shernan SK, Body SC. Utility of intraoperative transesophageal echocardiography for diagnosis of pulmonary embolism. Anesth Analg
14.Wareing TH, Davila-Roman VG, Barzilai B, et al. Management of the severely atherosclerotic ascending aorta during cardiac operations. A strategy for detection and treatment. J Thorac Cardiovasc Surg
15.Duda AM, Letwin LB, Sutter FP, et al. Does routine use of epiaortic ultrasonography decrease the stroke rate in coronary bypass surgery? J Vasc Surg
16.Rosenberger P, Shernan SK, Loffler M, et al. The influence of epiaortic ultrasonography on intraoperative surgical management in 6051 cardiac surgical patients. Ann Thorac Surg
17.Puskas JD, Williams WH, Duke PG, et al. Off-pump coronary artery bypass grafting provides complete revascularization with reduced myocardial injury, transfusion requirements, and length of stay: a prospective, randomized comparison of two hundred unselected patients undergoing off-pump verses conventional coronary artery bypass grafting. J Thorac Cardiovasc Surg
18.Katz ES, Tunick PA, Rusinek H, et al. Protruding aortic atheromas predict stroke in elderly patients undergoing cardiopulmonary bypass: experience with intraoperative transesophageal echocardiography. J Am Coll Cardiol
19.Molenberghs G, Kenward MG. Missing Data in Clinical Studies
. 1st ed. New York: John Wiley and Sons; 2007:105–117.
20.Blackstone EH. Comparing apples and oranges. J Thorac and Cardiovasc Surg
21.Ricci M, Karamanoukian HL, Abraham R, et al. Stroke in octogenarians undergoing coronary artery surgery with and without cardiopulmonary bypass. Ann Thorac Surg
22.Trehan N, Mishra M, Sharma OP, et al. Further reduction in stroke after off pump coronary artery bypass grafting: a 10-year experience. Ann Thorac Surg
23.Sharony R, Bizekis CS, Kanchuger M, et al. Off-pump coronary artery bypass reduces mortality and stroke in patients with atheromatous aortas: a case control study. Circulation
. 2003;108(suppl 1):II15–II20.
24.Ivanov J, Borger MA, Tu JV, et al. Mid-term outcomes of off-pump verses on-pump coronary artery bypass graft surgery. Can J Cardiology
25.Gold JP, Torres KE, Maldarelli W, et al. Improving outcomes in coronary surgery: the impact of echo-directed aortic cannulation and peri-operative hemodynamic management in 500 patients. Ann Thorac Surg
26.Bonatti J, Nagele G, Hangler H, et al. Extra-anatomical coronary artery bypass grafts on the beating heart of management of the severely atherosclerotic ascending aorta. Heart Surg Forum
. 2002;5(Suppl 4):S272–S281.
27.Kapetanakis EI, Stamou SC, Dullum MK et al. The impact of aortic manipulation on neurological outcomes after coronary artery bypass surgery: a risk-adjusted study. Ann Thorac Surg
28.Lev-Ran O, Braunstein R, Sharony R, et al. No-touch aorta off-pump coronary surgery: the effect on stroke. J Thorac Cardiovasc Surg
29.Vallely MP, Potger K, McMillan D, et al. Anaortic techniques reduce neurological morbidity after off-pump coronary artery bypass surgery. Heart Lung Circ
30.Zingone B, Rauber E, Gatti G, et al. Diagnosis and management of severe atherosclerosis of the ascending aorta and aortic arch during cardiac surgery: focus on aortic replacement. Eur J Cardiothorac Surg
31.Barbut D, Hinton RB, Szatrowski TP, et al. Cerebral emboli detected during bypass surgery are associated with clamp removal. Stroke
32.Calafiore A, Di Mauro M, Teodori G, et al. Impact of aortic manipulation on incidence of cerebrovascular accidents after surgical myocardial revascularization. Ann Thorac Surg
33.Gaspar M, Laufer G, Bonatti J, et al. Epiaortic ultrasound and intraoperative transesophageal echocardiography for the thoracic aorta atherosclerosis assessment in patient undergoing CABG. Surgical technique modification to avoid cerebral stroke. Chirurgia
34.Davila-Roman VG, Murphy SF, Nickerson NJ, et al. Atherosclerosis of the ascending aorta is an independent predictor of long term neurological events and mortality. J Am Coll Cardiol
35.Cahalan MK, Abel M, Goldman M, et al. American Society of Echocardiography and Society of Cardiovascular Anesthesiologists task force guidelines for training in perioperative echocardiography. Anesth Analg
This important study from Emory University emphasizes the importance and clinical utility of intraoperative epiaortic scanning in modern cardiac surgery. The authors reported the results in over 4000 patients undergoing EU at the time of coronary bypass grafting. The use of this technology altered aortic manipulation in more than 12% of patients. Patients with high-grade aortas had similar rates of death, stroke, or major adverse cardiac events when compared with patients with low grade aortas. The authors concluded that the use of this simple technology can minimize neurologic complications and allow surgeons to avoid excessive manipulation of diseased aortas. The examination takes only a few minutes to perform and can yield important information. It is recommended that EU be used liberally in cardiac surgery and should be mandatory in all patients with evidence of thoracic aortic atherosclerosis.