Endovascular repair of the aorta (EVAR) has steadily gained popularity as a reliable alternative to conventional surgical repair of aortic aneurysms since it was introduced into clinical practice in 1990 (1). The endovascular technique has also been successfully used to treat thoracic aortic diseases, offering a new option to patients considered “high risk” for open repair (2,3). The success of endovascular repair is critically dependent on demonstration of satisfactory graft deployment by various imaging modalities. Angiography is routinely performed during graft deployment to ensure accurate positioning and confirm the absence of perigraft endoleaks; classification of endoleaks is shown in Table 1(4). Transesophageal echocardiography (TEE) is also an ideal imaging tool for the thoracic and upper abdominal aorta and can supplement angiography in guiding placement of the endograft (5–7).
The proximity of the esophagus to the aorta in the intrathoracic and upper abdominal regions makes TEE an attractive imaging modality for aortic diseases. The sensitivity and specificity of TEE in the diagnosis of aortic pathology is well known (5). During EVAR, either in the operating room or the interventional radiology suite, TEE is the most sensitive imaging modality currently available for diagnosing endoleaks immediately after endograft deployment (8). Thoracic EVAR is accompanied by more hemodynamic fluctuations than abdominal EVAR as a result of more proximal aortic occlusion (9). The advantage of TEE is that in addition to imaging the descending thoracic aorta, it can also provide valuable information regarding cardiac function during endograft deployment. The anesthesiologist is ideally positioned to use intraoperative TEE to visualize the thoracic and upper abdominal aorta and monitor cardiac performance during thoracic EVAR. Although the utility of TEE in the endovascular setting has been reported (6–12), there is no similar published study in the anesthesiology literature. Therefore, we reviewed our experience of thoracic EVAR and the role of TEE in influencing intraoperative surgical decisions.
After approval by the Duke University Medical Center IRB, we reviewed data on all endovascular repairs of the thoracic or thoracoabdominal aorta performed at Duke University Medical Center. Seven patients underwent attempted thoracic EVAR in a 7-mo period between May, 2002 and February, 2003. Data were obtained from electronic medical records (from the Duke Medical Center Information Systems) and from the Saturn Anesthesia Information Management System (Drager, Telford, PA).
All patients were administered general anesthesia. Monitoring devices were inserted under local anesthesia in the induction room. These included a right radial arterial catheter, a pulmonary artery catheter with pulsed thermodilution continuous cardiac output monitoring capability (only in patients 1, 2, 3, and 6; the remaining had only a double-lumen central venous catheter placed), and a five-lead electrocardiogram. General endotracheal anesthesia was induced with fentanyl, midazolam and propofol. Anesthesia was maintained with isoflurane in an oxygen-air mixture. Muscular relaxation was achieved with vecuronium bromide.
After induction of anesthesia, an omniplane, phased array TEE probe (T6210 Omniplane II transducer; Phillips Medical Systems, Andover, MA) was introduced in all patients. Images were digitally acquired on a Phillips Sonos 5500 Ultrasound Imaging System (Phillips Medical Systems). In all patients, a comprehensive TEE examination was performed according to prescribed guidelines (13) after induction of anesthesia (at baseline) and after completion of the EVAR procedure. In addition to standard two-dimensional TEE images, color Doppler flow recordings were made for all valves. Left ventricular diastolic function was also noted by recording transmitral and pulmonary venous pulsed-wave Doppler flow velocities. As part of each TEE examination, the thoracic and upper abdominal aorta was scanned for the primary disease and coexisting pathology (e.g., additional aneurysms, atherosclerotic plaques). The thoracic aorta was imaged using the following views:
- Mid-esophageal ascending aortic long axis view at 120–150°.
- Mid-esophageal descending aorta long axis view at 90–110°.
- Mid-esophageal descending aorta short axis view at 0°.
- Upper esophageal aortic arch long axis view at 0°.
- Upper esophageal aortic arch short axis view at 90°.
During endograft deployment, TEE was used to visualize placement of the endograft in the thoracic aorta. After deployment the same views were used to visualize the placement of the endograft and check for endoleaks. In certain situations when the aneurysm was located at the junction of the arch and descending aorta, additional views were obtained with the transducer angles between 50°–70° at positions 4 and 5 above. Doppler color flow was also used to assess any flow within the aneurysmal sac. Because of the possibility of low-flow endoleaks, the aliasing velocity for Doppler color flow was reduced to 20–30 cm/s.
After ensuring accurate sizing of the aorta with preoperative imaging studies, the appropriate endograft was custom-built for each patient. IRB approval was obtained in each case for the use of a thoracic endograft. In most cases, the Talent Endoprosthesis (MedtronicAVE, Santa Rosa, CA) was planned for deployment. This device is a polyester fabric graft supported with a nitinol wire stent in a self-expanding configuration and is custom made up to 44 mm to treat up to 40-mm size aorta. It is covered with an integral sheath and has a diameter of 24F.
The appropriate “landing zone” for the endograft was determined by preoperative cross-sectional imaging and confirmed by intraoperative angiography. The endograft was inserted from the directly exposed femoral or iliac artery and advanced over a stiff guidewire to the thoracic position. Once the endograft stent system was positioned in the appropriate “landing zone,” mean arterial blood pressure was reduced to a level between 45 and 55 mm Hg with IV sodium nitroprusside. Heart rate was also decreased to a rate of 35–45 bpm with esmolol (bolus, 0.5 mg/kg). The sheath was then retracted to expose the endograft that subsequently self-expanded to coapt with the vessel wall. The graft was “annealed” or molded to the vessel wall with a compliant balloon to promote a tight seal. Immediately after deployment and balloon deflation the arterial blood pressure was increased to predeployment levels with discontinuation of vasodilator infusion and administration of phenylephrine in 100-μg increments if required. No maneuver was performed to increase the heart rate. On completion of endograft deployment, Doppler color flow mapping was performed at sites around the endograft to determine the presence of any endoleaks (see Table 1). Finally, after hemodynamic stabilization, a comprehensive TEE examination was once again performed to compare cardiac function with baseline values. All images were obtained and interpreted in real time by anesthesiologists certified by the US National Board of Echocardiography in perioperative transesophageal echocardiography (MS, JPM). Intraoperative TEE surveillance was performed by a dedicated anesthesiologist as well as by the primary anesthesia providers.
After completion of the procedure, reversal of relaxation and extubation was planned unless otherwise contraindicated. Patients were transferred to the postanesthesia care unit for observation. After meeting standardized discharge criteria, patients were transferred to a step-down unit until discharge from hospital. Routine follow-up protocol at our institution includes contrast-enhanced spiral computed tomography (CT) scans at 1, 6, and 12 mo to assess adequacy of aneurysmal sac exclusion.
The demographic profile of the six patients who underwent attempted thoracic EVAR is shown in Table 2. Successful endovascular repair was completed in six patients. One patient (#6) with type B aortic dissection underwent an attempted repair. Intraoperative TEE showed extensive dissection with the true lumen compressed by the false lumen throughout the length of the thoracic aorta. The TEE images also revealed spontaneous echo contrast (“echocardiographic smoke”) within the false lumen. The vascular surgery and radiology teams then determined that an endovascular repair could not be safely performed and that the patient was at high risk of aortic rupture. Based on TEE findings, the patient was scheduled for open repair by the cardiothoracic surgeons.
In five patients, the endograft stent system was passed through a femoral artery incision into the thoracic aorta for deployment. One patient (#1) required access through the distal abdominal aorta because of extensive trauma involving both femoral arteries. This patient also had concomitant aortofemoral bypass procedure after the EVAR. Because the endograft placement resulted in compromised circulation in the left arm (as it partially covered the origin of the left subclavian artery), a left carotid-subclavian artery bypass was also performed. Postoperative angiography in the operating room confirmed adequate flow in all areas. Patient #4 also underwent a concomitant aortofemoral bypass procedure because of a complication of EVAR, as detailed below (see Complications). Five patients received a custom-built Talent Endoprosthesis (Medtronic AVE) inserted, and one patient (#5) received an Aneurx Stent Graft System (Medtronic AVE). All six endografts were deployed successfully.
The aorta was well visualized in all patients until the descending aorta in the upper abdominal region (the nomenclature for this deep transgastric view of the descending aorta is not described in the American Society of Echocardiography/Society of Cardiovascular Anesthesiologists guidelines). Part of the aortic arch was obscured by the trachea and could not be evaluated. The primary aortic disease was accurately identified in all patients. All patients had varying degrees of aortic atherosclerosis but no significant aortic pathology other than the primary disease. At the start of the EVAR procedure, the guidewire was visualized and confirmed in the true aortic lumen. The TEE probe also served as a useful marker of the location of the aneurysm on fluoroscopy. After initial angiography the aneurysm was identified by filling of the sac with contrast dye. The TEE probe then served as a useful marker (as long as it was not moved excessively during this period) of the correct location when subsequent noncontrast fluoroscopy was performed to advance the endograft to the aneurysm location. The stent graft system was also seen by TEE in the thoracic aortic lumen before deployment (Fig. 1). The left subclavian artery was identified accurately in all patients. In patient #1, this artery was partially occluded by the graft. This flow obstruction was identified with both angiography and TEE intraoperatively. After endograft deployment both proximal and distal limits of the stent were accurately identified by TEE.
Endoleaks of different types (see Table 1 for a description of the current consensus on endoleak terminology) were also observed in a number of patients. In one patient (#3), an endoleak was observed by TEE as color flow within the aneurysmal sac (see Fig. 2 and video loop 1 online at http://www.anesthesia-analgesia.org/). Subsequent angiography in this patient was unable to confirm the endoleak. Based on TEE findings alone, remedial action was undertaken by deploying a second stent to obliterate the endoleak. The endoleak was still observed by TEE but not by angiography. Initially, this was suspected as a type I endoleak, but lack of evidence of an entry site of flow to the aneurysmal sac by TEE prompted speculation to the possibility of a Type IIA or IIB leak. Postoperative followup spiral tomography in this patient 3 wk later revealed a persistent endoleak in the mid-region of the endograft, without increase in size of the aneurysmal sac. However, at 3-mo followup, spiral CT scan showed that there was no endoleak, indicating spontaneous exclusion of the aneurysmal sac and reinforcing the possibility that it was a type II endoleak. In patient #4, a type 1B endoleak from the distal attachment site of the endograft was observed by TEE (video loop 2). Subsequent angiography confirmed this leak. Remedial action was undertaken in the form of deployment of an extension cuff beyond the distal end of the original endograft, and the leak was sealed off from the aneurysmal sac. Table 3 provides a description of the types of endoleaks seen in our case series and the modality by which they were detected.
In patient #2, a small type IIA endoleak was observed as color flow from a branch vessel into the aneurysmal sac (video loop 3). This flow was deemed to be nonsignificant in terms of ability to enlarge aneurysm size. Postoperative followup CT scans confirmed the absence of any endoleak and a stable aneurysmal sac size.
There were no TEE-related complications (e.g., esophageal injury, bleeding) in any patient. One patient (#4) had an avulsion of an existing aortobifemoral prosthesis during EVAR. As the hemorrhagic complication occurred within a closed space, not visible to the surgical team, it was not suspected until acute hypovolemia was identified by TEE as the reason for sudden hypotension. Immediate control of bleeding and hemodynamic status was achieved by inflating the aortic balloon proximal to the avulsion site and performing an open revision of the aortobifemoral bypass. TEE was also helpful in guiding resuscitation efforts in this patient.
TEE is a useful imaging modality during endovascular repair of the thoracic aorta. It is the most sensitive tool for detecting endoleak in the immediate postdeployment period, and it supplements angiography in comprehensively imaging the aorta during EVAR (6–8). The information derived from TEE may also lead to modifications in endograft positioning or repositioning and indicate whether EVAR is a feasible treatment option.
We were able to identify endoleaks by TEE in three patients immediately after endograft deployment. Endoleaks, defined as persistent blood flow outside the endograft and within the aneurysm sac, may occur in nearly 20% of EVARs for abdominal aortic aneurysms (14). No clinical trials have compared the incidence of endoleaks between thoracic and abdominal EVARs. The initial evaluation of endoleaks around endografts in the thoracic position revealed an incidence of 24%(3). Recent studies of thoracic EVAR report the incidence of endoleaks between 20% and 28%(15–17). Most centers, including ours, evaluate endoleaks after abdominal aortic aneurysms with intraoperative angiography, intravascular ultrasound, or postoperative CT scanning. In contrast, endoleaks after thoracic EVAR may be more readily diagnosed by TEE, leading to a more frequent reported incidence in the literature. However, the use of TEE for the diagnosis of endoleaks is not consistent across all centers performing thoracic EVAR. The currently more frequent reported incidence of endoleaks after thoracic EVAR may be possibly the result of the higher intraluminal pressure within the thoracic aorta and improved diagnosis of endoleaks by TEE. Endoleaks may be found on postdeployment imaging or on followup CT scans several weeks after the EVAR. The advantage of detecting an endoleak in the operating room is that an immediate decision may be made to repair the leak by repositioning the endograft or by deploying a second prosthesis at the endoleak site. Accurate imaging is, therefore, crucial if this common complication of EVAR is to be avoided.
The sensitivity of TEE in detecting small amounts of flow within the aneurysmal sac may make it ideal for the detection of type I and II endoleaks. In many instances, TEE is frequently able to detect an endoleak that cannot be confirmed by angiography (6–8). Doppler color flow mapping is a sensitive technique for assessment of blood flow in any area. Barbetseas et al. (18) were able to diagnose ascending aortic pseudoaneurysms with Doppler echocardiography in seven of eight patients. Three of these were not seen by aortography. Rapezzi et al. (6) found that in seven patients with endoleak after thoracic EVAR, TEE was accurate in all seven, but angiography failed to diagnose the problem in five of these cases. Another recent study of 12 EVARs for acute thoracic type B dissections found intraoperative TEE more sensitive than angiography in three patients (7). In a study of 25 patients undergoing thoracic EVAR, Fattori et al. (8) reported that six cases of endoleak detected by intraoperative TEE were missed by angiography. Additionally, these investigators also reported that TEE was valuable as a guide to the right aortic lumen. The sensitivity of TEE in diagnosing endoleaks has been attributed to its ability to image the space between the endograft and aortic wall. The Doppler color flow modality confers the additional advantage of detecting flow in this area. The disadvantage of angiography in this setting is that it relies on a fixed volume of dye to circulate within the endoleak. Small leaks may be missed because the volume of dye within the leak may not be detectable by fluoroscopy or because the imaging angle may not be sufficiently accurate to detect the leak. Another advantage of TEE is that, unlike angiography, its image acquisition capability is not dependent on potentially nephrotoxic contrast dye. In our review, we found TEE to be a reliable indicator of endoleak. However, the question then arises: when an endoleak is detected by TEE but not by angiography in the operating suite, is it significant?
When an endoleak is detected by both angiography and TEE, it will probably be deemed significant, and remedial action will be taken to seal it by placing a second stent/extension cuff. However, a problem occurs when the endoleak is seen only by TEE but not by angiography. TEE may be sensitive enough to detect very small endoleaks that escape detection by angiography. These endoleaks may or may not be significant enough to impact on short- or long-term outcome. In our case (patient #3), the surgeons preferred to take the safer option by attempting to seal the TEE-detected endoleak. However, attempts at re-sealing failed and the endoleak was subsequently detected by spiral CT scan 3 weeks later. However, the endoleak proved to be insignificant, as it had spontaneously closed off after 3 months. Only prospective observational trials that link endoleak detection by standardized methods with outcome variables will be able to determine whether remedial action should be taken in the operating room based on TEE findings alone. Currently no such data exist, and this scenario remains a diagnostic dilemma. At the present time, we cannot recommend that remedial action be taken each time an endoleak is detected by TEE alone. Undertaking remedial action is a major endeavor that involves more contrast dye, further manipulation and instrumentation of the aorta, increase in operating time and expense, and potentially more risk. This decision must be based on the clinical circumstances, TEE evidence, and consensus among surgeons, radiologists, and anesthesiologists in each individual case. Experience with endoleak management and postoperative followup CT scans will help determine which TEE-detected endoleaks are clinically significant.
The timely detection of endoleaks may be one of the most important benefits of TEE during thoracic EVAR. The principal advantage of TEE is that it enables the anesthesiologist to identify aortic pathology, detect endoleaks, and monitor cardiac performance with a single imaging tool. We also found TEE invaluable in providing real-time information about guidewire position within the true lumen in cases of aortic dissection, a finding that has been previously reported (6).
Myocardial responses to aortic cross-clamping are well known. The higher the level of the clamp, the greater is the hemodynamic disturbance. Unlike open aortic aneurysm repair, endovascular techniques do not involve extended periods of aortic occlusion. During thoracic EVAR, significant hemodynamic manipulation is required to ensure accurate graft placement. A significant advantage of TEE is that it can effectively determine whether myocardial function has been compromised as a result of hemodynamic disturbances.
Our report is limited by the small cohort of patients evaluated. However, this is an important new development, and small reports such as this one will provide information to anesthesiologists and form the basis of prospective large-scale clinical trials. With regard to detection of endoleaks, although TEE is a sensitive tool for the measurement of low velocity flow, it may not be able to detect the difference between actual flow within the sac or movement of blood because of transmitted pulsations. The use of different imaging planes and pulsed wave Doppler may help identify endoleaks more accurately and overcome this limitation. A limitation of TEE is that it is unable to visualize more distal abdominal aneurysms. In cases where a thoracoabdominal aortic aneurysm extends into the distal abdominal aorta (Crawford Extent II), TEE can only visualize the thoracic and upper abdominal components of the aneurysm. The portion below the origin of the celiac trunk may be lost to view because the TEE probe cannot be inserted any further than the stomach. However, most of these extensive thoracoabdominal aneurysms are not amenable to EVAR and must undergo an open repair because of the length of aortic involvement. Most thoracic EVARs involve only the distal aortic arch or descending thoracic aorta. Each of these areas is well visualized by TEE. Although imaging was excellent for all our patients, the only area where difficulty was encountered was in the proximal aortic arch, which is usually not visualized by TEE because of the interposition of the trachea between the esophagus and the aorta at this level (19). Although this is a problem for endografts that extend more proximally, this was not a concern in our EVAR patients because at our institution aneurysms involving the ascending aorta and arch are managed by the cardiac surgical team as an open repair with cardiopulmonary bypass. However, in institutions that do perform EVARs for more proximally extending aneurysms, TEE may be insufficient for evaluation of Type IA endoleaks.
Although this subject has been previously studied, it has been reported almost exclusively in surgery or radiology journals. There has been only one publication in the anesthesiology literature that has dealt with the role of TEE in thoracic EVAR. This was a review article on anesthesia for endovascular repairs in general that had a section that reviewed some aspects of the use of TEE in this patient population (20). However, there has been no study in anesthesiology literature that has highlighted the role of TEE in thoracic EVAR. Most publications on this subject have been authored by surgeons, cardiologists, or radiologists reporting in nonanesthesia journals. Anesthesiologists have a pivotal role in influencing outcomes by using a valuable imaging tool in the operating room. Unfortunately, anesthesiologists are not publishing their findings on this topic in journals that have the most readerships among anesthesia providers. This report will provide a large audience of anesthesia providers with insight into the importance of intraoperative TEE in thoracic EVARs.
In summary, we found TEE to be an invaluable imaging tool during endovascular repair of aortic pathology. The importance of TEE in thoracic EVAR is slowly emerging (6,8,11,12). As with any new technique, the utility of different supportive techniques such as TEE will be better known only with experience. TEE can supplement information obtained by angiography to enhance the accuracy of EVAR and potentially improve outcomes. The anesthesiologist is in an ideal position to provide the endovascular team with vital information regarding stent positioning, endoleaks, and cardiac performance with a single imaging modality.
The authors would like to gratefully acknowledge the contribution of Ms. Dheadra McAdoo, Echocardiography Technician, Intraoperative TEE Service, Division of Cardiothoracic Anesthesiology, Duke University Medical Center.
1. Parodi JC, Palmaz JC, Barone HD. Transfemoral intraluminal graft implantation for abdominal aortic aneurysms. Ann Vasc Surg 1991; 5: 491–9.
2. Dake MD, Miller DC, Semba CP, et al. Transluminal placement of endovascular stent-grafts for the treatment of descending thoracic aortic aneurysms. N Engl J Med 1994; 331: 1729–34.
3. Mitchell RS, Miller DC, Dake MD, et al. Thoracic aortic aneurysm repair with an endovascular stent graft: the “first generation.” Ann Thorac Surg 1999; 67: 1971–4.
4. Veith FJ, Baum RA, Ohki T, et al. Nature and significance of endoleaks and endotension: summary of opinions expressed at an international conference. J Vasc Surg 2002; 35: 1029–35.
5. Yalcin F, Thomas JD, Homa D, Flachskampf FA. Transesophageal echocardiography: first-line imaging for aortic diseases. Cleve Clin J Med 2000; 67: 417–8, 21–8.
6. Rapezzi C, Rocchi G, Fattori R, et al. Usefulness of transesophageal echocardiographic monitoring to improve the outcome of stent-graft treatment of thoracic aortic aneurysms. Am J Cardiol 2001; 87: 315–9.
7. Gonzalez-Fajardo JA, Gutierrez V, San Roman JA, et al. Utility of intraoperative transesophageal echocardiography during endovascular stent-graft repair of acute thoracic aortic dissection. Ann Vasc Surg 2002; 16: 297–303.
8. Fattori R, Caldarera I, Rapezzi C, et al. Primary endoleakage in endovascular treatment of the thoracic aorta: importance of intraoperative transesophageal echocardiography. J Thorac Cardiovasc Surg 2000; 120: 490–5.
9. Kahn RA, Moskowitz DM, Marin M, Hollier L. Anesthetic considerations for endovascular aortic repair. Mt Sinai J Med 2002; 69: 57–67.
10. Kahn RA, Konstadt S. Thoracic aortic disease: endovascular stents. Echocardiography 2002; 19: 589–97.
11. Orihashi K, Matsuura Y, Sueda T, et al. Echocardiography-assisted surgery in transaortic endovascular stent grafting: role of transesophageal echocardiography. J Thorac Cardiovasc Surg 2000; 120: 672–8.
12. Moskowitz DM, Kahn RA, Konstadt SN, et al. Intraoperative transoesophageal echocardiography as an adjuvant to fluoroscopy during endovascular thoracic aortic repair. Eur J Vasc Endovasc Surg 1999; 17: 22–7.
13. Shanewise JS, Cheung AT, Aronson S, et al. ASE/SCA guidelines for performing a comprehensive intraoperative multiplane transesophageal echocardiography examination: recommendations of the American Society of Echocardiography Council for Intraoperative Echocardiography and the Society of Cardiovascular Anesthesiologists Task Force for Certification in Perioperative Transesophageal Echocardiography. Anesth Analg 1999; 89: 870–84.
14. van Marrewijk C, Buth J, Harris PL, et al. Significance of endoleaks after endovascular repair of abdominal aortic aneurysms: the EUROSTAR experience. J Vasc Surg 2002; 35: 461–73.
15. Pamler RS, Kotsis T, Gorich J, et al. Complications after endovascular repair of type B aortic dissection. J Endovasc Ther 2002; 9: 822–8.
16. Schoder M, Cartes-Zumelzu F, Grabenwoger M, et al. Elective endovascular stent-graft repair of atherosclerotic thoracic aortic aneurysms: clinical results and midterm follow-up. AJR Am J Roentgenol 2003; 180: 709–15.
17. Orend KH, Scharrer-Pamler R, Kapfer X, et al. Endovascular treatment in diseases of the descending thoracic aorta: 6-year results of a single center. J Vasc Surg 2003; 37: 91–9.
18. Barbetseas J, Crawford ES, Safi HJ, et al. Doppler echocardiographic evaluation of pseudoaneurysms complicating composite grafts of the ascending aorta. Circulation 1992; 85: 212–22.
19. Konstadt SN, Reich DL, Quintana C, Levy M. The ascending aorta: how much does transesophageal echocardiography see? Anesth Analg 1994; 78: 240–4.
© 2003 International Anesthesia Research Society
20. Kahn RA, Moskowitz DM. Endovascular aortic repair. J Cardiothorac Vasc Anesth 2002; 16: 218–33.