CARDIOVASCULAR
The Reversed Elephant Trunk Technique Used for Treatment of Complex Aneurysms of the Entire Thoracic Aorta Joseph S. Coselli, MD, Scott A. LeMaire, MD, Stacey A. Carter, BA, and Lori D. Conklin, MD The Texas Heart Institute at St. Luke’s Episcopal Hospital and the Division of Cardiothoracic Surgery, Baylor College of Medicine, Houston, Texas
Objective. A preferred technique for the staged treatment of patients with aneurysms involving the entire thoracic aorta is the elephant trunk technique, with replacement of the proximal (ascending and transverse aortic arch) aorta as the initial procedure. Some patients, however, need to have the distal aortic segments (descending and thoracoabdominal aorta) addressed during the first operation. We evaluated outcomes in a series of patients who underwent distal aortic replacement first using the reversed elephant trunk technique. Methods. Thirty-eight patients underwent first-stage graft repair of the descending thoracic (n ⴝ 3) or thoracoabdominal (n ⴝ 35) aorta using the reversed elephant trunk technique. Twelve patients (32%) ultimately underwent second-stage aortic arch replacement after a mean interval of 3.9 months (range, 1.6 –14 months).
Results. The operative mortality for the initial procedure was 16% (6/38 patients). One patient had a stroke (3%) and 1 patient developed paraparesis (3%). In the interval between the 2 procedures, there were 4 late deaths (4/32; 13%), 1 due to respiratory failure and 3 due to unknown causes. After the 12 completion procedures, there was 1 in-hospital death (8%) and there were no strokes. Five-year survival for the overall group was 51.3 ⴞ 10.8%. Conclusions. Surgical treatment of aneurysms involving the entire thoracic aorta remains challenging and is associated with substantial morbidity and mortality. The reversed elephant trunk technique facilitates staged repair in patients who require distal aortic replacement during the first operation. (Ann Thorac Surg 2005;80:2166 –72) © 2005 by The Society of Thoracic Surgeons
Introduction
distal anastomosis where adhesions may place the esophagus, vagus nerve, recurrent laryngeal nerve and pulmonary artery at risk. In a subset of patients with extensive thoracic aortic aneurysms, the descending thoracic or thoracoabdominal segment is symptomatic or disproportionately large compared to the ascending/arch aorta, necessitating distal repair as the initial procedure. To facilitate staged repair in such patients, Carrel and Althaus applied the elephant trunk principle to descending thoracic aortic replacement; in this “reversed elephant trunk technique,” the graft is folded within itself (invaginated) in the proximal portion of the descending aorta immediately distal to the left subclavian artery [7]. During the subsequent operation, the elephant trunk segment is used to replace the transverse aortic arch, obviating the need for a distal anastomosis under circulatory arrest. The purpose of this report is to describe our experience using the reversed elephant trunk technique in 38 patients.
Aneurysmal disease of the thoracic aorta can present as an extensive process involving both the proximal and distal segments (Fig 1) [1]. The management of these extensive aortic aneurysms continues to pose a challenge. Although graft replacement of the ascending aorta, transverse aortic arch, and descending or thoracoabdominal aorta as a single procedure has been described, its application has been limited, partially as a consequence of the substantial associated morbidity and mortality [2– 4]. An ingenious approach to the staged repair of extensive thoracic aortic aneurysms was the procedure originally described by Borst and colleagues and frequently referred to as the elephant trunk technique [5– 6]. In this approach, a free-floating portion of the vascular prosthesis is positioned within the descending thoracic aorta at the time of ascending and aortic arch replacement. The intent is to facilitate the second procedure and reduce the risk of distal aortic reconstruction by eliminating the need for extensive dissection at the site of the previous Accepted for publication March 21, 2005.
Patients and Methods
Presented at the Fiftieth Annual Meeting of the Southern Thoracic Surgical Association, Bonita Springs, FL, Nov 13–15, 2003.
Patients
Address correspondence to Dr Coselli, Division of Cardiothoracic Surgery, Baylor College of Medicine, One Baylor Plaza, BCM 390, Houston, TX 77030; e-mail:
[email protected].
We began using the reversed elephant trunk technique in 1994. Since that time, we have carried out 1,878 operations on the descending thoracic (n ⫽ 293) or
© 2005 by The Society of Thoracic Surgeons Published by Elsevier Inc
0003-4975/05/$30.00 doi:10.1016/j.athoracsur.2005.03.087
Figure 1. Drawing of an aneurysm involving the ascending aorta, transverse aortic arch and thoracoabdominal aorta.
thoracoabdominal aorta (n ⫽ 1,585). Of these, 38 patients (2.0%) underwent replacement of the distal thoracic aorta with the reversed elephant trunk technique. Patients with anterior chest pain, extensive critical coronary artery disease, or severe valvular pathology were excluded from this approach and required an initial proximal repair via median sternotomy to enable treatment of valvular and cardiac issues. The characteristics of the 38 patients are presented in Table 1. Clinical data was prospectively entered into a computerized database beginning at the time of initial evaluation. Forty-five percent of patients were symptomatic; the most common symptom was back pain. Emergent or urgent operations were required in 13 patients (34%) with acute presentation, including 3 (8%) with contained rupture. One patient (3%) presented with acute dissection superimposed upon a preexisting degenerative aneurysm and was operated upon emergently. Fifty percent of patients had previously undergone aortic operations, including 12 proximal, 1 descending thoracic, 1 Crawford extent IV thoracoabdominal, and 5 infrarenal abdominal aortic repairs.
Distal Aortic Repairs (Stage 1) The initial operations in the 38 patients were 3 (8%) descending thoracic repairs, 14 (37%) Crawford extent I thoracoabdominal repairs, and 21 (55%) Crawford extent II thoracoabdominal repairs. Our approach to graft replacement of descending thoracic and thoracoabdominal aortic aneurysms has been previously reported [8 –11]. The aneurysms were exposed through a standard left posterolateral thoracotomy or a thoracoabdominal incision through the 6th intercostal space. For all thoracoabdominal repairs, the diaphragm was divided in a circular fashion, leaving a rim of 4 cm on the chest wall for later reattachment. The abdominal viscera were rotated medi-
COSELLI ET AL REVERSED ELEPHANT TRUNK TECHNIQUE
2167
ally allowing exposure of the entire descending thoracic and abdominal aorta. In 27 patients (71%), left heart bypass with a centrifugal pump was employed to provide distal aortic perfusion during the proximal anastomosis. In 23 patients (61%), cerebrospinal fluid drainage was utilized. The aorta was controlled proximally either immediately distal to the left subclavian artery or between the left common carotid and the left subclavian arteries. When left heart bypass was used, a distal clamp was placed across the mid-descending thoracic aorta. After opening the proximal portion of the aneurysm, if dissection was present, the wall between the true and false lumens was removed. Proximal intercostal arteries were oversewn if present and backbleeding vigorously. The proximal descending thoracic aorta was transected and separated from the esophagus, taking care to avoid injury to the vagus and recurrent laryngeal nerves. An appropriately sized polyester graft was selected and a portion approximately 8 cm in length was invaginated to create the reversed elephant trunk (Fig 2). An end-to-end anastomosis was performed between the aorta and the folded edge of the graft using continuous polypropylene suture. If a clamp had been placed proximal to the left subclavian artery, the clamp was moved down onto the graft after the proximal anastomosis was completed to restore flow to the left subclavian artery (Fig 3). At this point, left heart bypass was discontinued, the aneurysm was opened to its distal extent, and selective celiac and superior mesenteric artery perfusion was started. Cold crystalloid soluTable 1. Preoperative Characteristics of 38 Patients Who Underwent Descending Thoracic or Thoracoabdominal Aortic Aneurysm Repair (Stage 1) Using the Reversed Elephant Trunk Technique Variable
No. of Patients (%)
Men Women Age (years) Mean Median Range Dissection Acute Chronic Non Symptomatic aneurysm Acute presentation Rupture Marfan syndrome Renal insufficiency Cerebrovascular disease Insulin dependent diabetes Coronary artery disease Hypertension Chronic pulmonary disease Prior coronary artery bypass Previous aneurysm repair
21 (55) 17 (45) 64.6 67.5 24–76 1 (3) 15 (39) 22 (58) 17 (45) 13 (34) 3 (8) 8 (21) 3 (8) 5 (13) 1 (3) 12 (32) 24 (63) 6 (16) 6 (16) 19 (50)
CARDIOVASCULAR
Ann Thorac Surg 2005;80:2166 –72
2168
COSELLI ET AL REVERSED ELEPHANT TRUNK TECHNIQUE
Ann Thorac Surg 2005;80:2166 –72
CARDIOVASCULAR Figure 2. The initial steps of the reversed elephant trunk technique: after aortic clamping, the aneurysm is opened, the graft is inverted upon itself (arrow in inset) and the folded edge is sutured end-toend to the proximal descending aorta.
tion was delivered to the right and left renal arteries for renal protection. Selected intercostal and lumbar arteries were reattached to a side-opening in the graft in 30 patients (79%). The visceral vessels were reattached individually or as a group, depending upon anatomic proximity. The clamp on the graft was often sequentially moved distally during the repair, restoring flow to reattached branch arteries. The distal anastomosis completed the first stage of the reconstruction (Fig 4). Median total aortic clamp time was 45 minutes.
Proximal Aortic Repairs (Stage 2) In 12 patients, the second stage was performed a mean interval of 3.9 months (range, 1.6 –14 months) after the initial procedure. All second-stage procedures were per-
Figure 3. After completion of the proximal anastomosis, the intercostal arteries have been reattached to an opening in the side of the graft. In this circumstance, the distal anastomosis is being constructed as a bevel at the level of the visceral and renal vessels.
Figure 4. Postoperative drawing illustrating the completed thoracoabdominal repair with the invaginated reversed elephant trunk graft suspended in the proximal portion of the graft.
formed via standard median sternotomy and included replacement of the ascending aorta and the entire transverse aortic arch. In the 3 patients who had preexisting ascending aortic grafts, the remaining segment of ascending aorta beyond the old graft was replaced. Intraoperative variables are listed in Table 2. Cardiopulmonary bypass was usually established via cannulas placed in the aneurysmal ascending aorta and both vena cavae. After achieving profound hypothermic circulatory arrest, the ascending portion of the aneurysm was opened longitudinally into the transverse aortic arch, again avoiding injury to the vagus nerve. Adjuncts for cerebral protection included retrograde cerebral perfusion in 3 patients and selective antegrade perfusion in 2 patients [12]. If aortic dissection was present, the dissecting membrane between the true and false lumens was excised. The invaginated graft was retrieved and pulled back into the field to facilitate the proximal repair (Fig 5). The brachiocephalic vessels were attached either directly (Fig 6) or by separate grafts, depending on anatomical considerations. One patient underwent concomitant innominate artery bypass and 1 patient underwent innominate, left common carotid and left subclavian artery bypass. After deairing, the graft was clamped, cardiopulmonary bypass was resumed (Figure 7), and rewarming was initiated. Concomitant procedures carried out during the proximal aortic reconstruction included coronary artery bypass grafting in 2 patients, aortic valve annuloplasty in 3 patients and aortic root replacement with a cryopreserved homograft in 1 patient. An example of the completed repair is illustrated in Figure 8.
Ann Thorac Surg 2005;80:2166 –72
COSELLI ET AL REVERSED ELEPHANT TRUNK TECHNIQUE
2169
Variable
No. of Patients (%)
Retrograde cerebral perfusion Antegrade cerebral perfusion Cooling time (min) Mean Median Range Rewarming time (min) Mean Median Range Circulatory arrest time (min) Mean Median Range Aortic clamp time (min) Mean Median Range Cardiopulmonary bypass time (min) Mean Median Range Concomitant procedures Innominate artery replacement Three branch arch graft Coronary artery bypass Aortic valve annuloplasty Aortic root replacement (homograft)
3 (25) 2 (17)
CARDIOVASCULAR
Table 2. Intraoperative Variables for 12 Patients Who Underwent Ascending/Aortic Arch Aneurysm Repair (Stage 2) for Completion of the Reversed Elephant Trunk Technique of Total Thoracic Aortic Replacement
33 38 25–40 86 84 69–100 39 39 27–48 54 43 21–97 120 124 94–137 1 (8) 1 (8) 2 (17) 3 (25) 1 (8)
Figure 5. The second stage of the reversed elephant trunk procedure involves opening the transverse aortic arch via median sternotomy under profound hypothermic circulatory arrest. The invaginated graft segment within the proximal descending thoracic aorta is withdrawn (arrow).
Interval between Stages Four of the 32 surviving patients (13%) died during follow-up, before the second stage of repair could be completed. One of the late deaths was due to respiratory
Follow-up Current follow-up was available in all 38 patients. Survival was estimated using the Kaplan-Meier method; survival curves were compared using the log-rank test.
Results Distal Aortic Repair (Stage 1) There were 6 operative deaths (16%), including 2 patients who died within 30 days of surgery (5%). All 6 early deaths were the result of multiple organ failure or sepsis. Pulmonary complications occurred in 13 patients (34%). There were no cases of paraplegia. Paraperesis developed in 1 patient (3%) who underwent emergent extent II thoracoabdominal aortic repair; this patient was ambulating with the assistance of a walker at discharge and to date is reportedly walking on his own. One patient suffered a stroke (3%), 2 patients (5%) required hemodialysis, and 4 patients had cardiac complications (2 with atrial fibrillation and 2 with myocardial infarction). Four patients (11%) developed left vocal cord paralysis.
Figure 6. The reversed elephant trunk eliminates the need for a distal anastomosis during the proximal repair; the brachiocephalic vessels are reattached to an opening in the side of the graft.
2170
COSELLI ET AL REVERSED ELEPHANT TRUNK TECHNIQUE
Ann Thorac Surg 2005;80:2166 –72
CARDIOVASCULAR Figure 9. Kaplan-Meier survival curve for the 38 patients who underwent the first-stage operation using the reversed elephant trunk technique.
(n ⫽ 3), or the patient does not want to undergo another major operation (n ⫽ 1). Figure 7. Cardiopulmonary bypass is reestablished and rewarming begun as the proximal anastomosis is carried at the sinotubular junction.
failure at 14.3 months following initial repair. The other 3 deaths occurred at 21.3, 32.8 and 45.4 months following initial repair and were due to unknown causes. To date, 16 surviving patients (50%) have not yet received a second-stage intervention. The reasons for not proceeding with the second operation include: the proximal aneurysm has not yet reached a size for which operative intervention is recommended (n ⫽ 12), postoperative complications from the initial procedure are prohibitive
Proximal Aortic Repair (Stage 2) Twelve of the remaining 28 patients (43%) underwent completion procedures after a mean interval of 3.9 months following initial repair. No patients died within 30 days of operation. There was 1 (8%) in-hospital death due to multiple organ failure 3.5 months following completion repair. No patients suffered stroke or paraplegia/ paraparesis, but 1 patient (8%) developed severe encephalopathy. Pulmonary complications occurred in 3 patients (25%). Four late deaths (33%) occurred after the second-stage repair, 1 due to multiple organ failure 5.3 months postoperatively and 3 due to unknown causes at 1.3, 34, and 96.5 months following the operation.
Mid-term Survival Cumulative 5-year survival for the entire group of 38 patients was 51.3 ⫾ 10.8% (Figure 9). Three-year survival rates were similar in patients that have undergone sec-
Figure 8. Postoperative drawing following completion of the reversed elephant trunk procedure for repair of the entire thoracic aorta.
Figure 10. Mid-term survival curves were similar in patients who underwent the second-stage repair (solid line) and patients who did not undergo the second-stage repair (dotted line; p ⫽ 0.59).
ond-stage repair (59.3 ⫾ 16.8%) and those who have not undergone second-stage repair (57.8 ⫾ 11.7%, p ⫽ 0.59; Figure 10).
Comment It is not uncommon for aneurysms involving 1 portion of the thoracic aorta to extend into contiguous segments. Replacement of the entire thoracic aorta in a single setting can be performed via a bilateral submammary anterior thoracotomy, median sternotomy with a lateral thoracotomy extension, or median sternotomy with a separate lateral thoracotomy or thoracoabdominal incision [2– 4]. While the single-stage approach eliminates the interval between operations and the attendant risk of rupture, we believe that these extensive procedures cause considerable morbidity – including respiratory problems, bleeding and neurological complications – particularly in elderly patients. Furthermore, this approach is primarily advocated for aneurysms that do not extend below the diaphragm; singlestage procedures are rarely employed for thoracoabdominal aortic aneurysms, which were present in over 90% of the patients in this series. In most patients with aneurysms involving the entire thoracic aorta, the ascending and transverse aortic arch can be treated during the initial procedure. In these patients, we prefer the elephant trunk technique described by Borst and associates [5,6], in part because this approach enables treatment of valvular and coronary artery occlusive disease during the first stage. After the traditional elephant trunk operation, the second-stage distal procedure is safer because there is no need for dissection near the distal arch anastomosis, thereby reducing the risk of injury to adjacent structures, such as the phrenic nerve, recurrent laryngeal nerve, pulmonary artery, and esophagus. Some patients with extensive thoracic aortic disease, however, have descending thoracic or thoracoabdominal aortic aneurysms that are disproportionately larger than the proximal segment, are symptomatic (usually manifested by back pain), have ruptured or demonstrate evidence of impending rupture on imaging studies. In such patients, the distal aortic segment is treated at the initial procedure, employing the reversed elephant trunk technique, followed by a second-stage repair of the ascending and transverse aortic arch. In 1997, Carrel and Althaus [7] described using this approach in 3 patients who required initial repair of the descending thoracic aorta. The following year, we reported using this method during extent I replacement of a thoracoabdominal aortic aneurysm prior to a second stage repair of the ascending and arch aorta [13]. Shiiya and colleagues [14] described 2 patients with Marfan syndrome who underwent the first stage of reversed elephant trunk reconstruction for thoracoabdominal aortic aneurysms caused by aortic dissection; at the time of their report, the secondstage procedure had not been carried out in either case. Zanetti [15] described the use of the reversed elephant trunk technique in a 63-year old man in whom the descending thoracic aorta was disproportionately large and symptomatic. The second stage was successfully carried out 4 months later. In 2001, Carrel and colleagues [16] reported
COSELLI ET AL REVERSED ELEPHANT TRUNK TECHNIQUE
2171
their experience with 17 patients who were treated with the reversed elephant trunk technique; 9 patients subsequently underwent second-stage reconstruction of the aortic arch, with the remainder waiting. The vast majority of previously reported cases involved repairs limited to the descending thoracic aorta; in contrast, in our series, only 8% of patients had descending thoracic aortic repairs while 92% of patients had more extensive thoracoabdominal aortic repairs. The reversed elephant trunk technique provides several advantages over single-stage repairs or non-elephant trunk approaches. During the first-stage distal aortic procedure, for example, this technique allows the proximal descending thoracic aortic anastomosis to be performed well beyond the level of the recurrent laryngeal and phrenic nerves, reducing the risk of injury to these structures. Retraction of the left lung during full heparinization is also avoided, potentially reducing pulmonary complications. During the second-stage arch operation, benefits include: elimination of the distal aortic anastomosis, reduction of the circulatory arrest time, improved exposure of the supraaortic vessels, and reduction of the risk of hemorrhage [16]. Tailoring the operation based upon a patient’s individual needs and anatomic considerations is crucial. The presentation of pathology is immensely varied and consequently, the specific technical aspects of each procedure are modified to address individuality. As a result, as represented in this series, the extent of the descending thoracic and thoracoabdominal aortic repair naturally varies. Further, the technique may be employed in patients in whom the transverse aortic arch or ascending aorta have not yet met criteria for graft replacement, but are likely to require replacement in the future. These patients are carefully followed with serial computed tomography scans and undergo the second-stage procedure once indications are present. This report must be interpreted in the context of the limitations inherent to all retrospective studies. Several aspects of perioperative management have changed substantially over the study period, resulting in considerable heterogeneity in treatment. While current follow-up data was available for all of the patients, the causes of most of the late deaths were unknown. Hence, while there were no known aortic ruptures during the interval between operations, we cannot comment on whether rupture contributed to the interval mortality in this series. Given the potential for aortic rupture after the first stage, we currently recommend that patients undergo the second operation within 4 to 6 weeks whenever possible. Surgical treatment of aneurysms involving the entire thoracic aorta remains challenging and is associated with substantial morbidity and mortality. Despite the high risks associated with operative management, the mid-term survival is rewarding in the context of the natural history of unrepaired aneurysms. In the report by Crawford and DeNatale [17], fewer than 20% of patients with unrepaired thoracoabdominal aortic aneurysms were alive 5 years after diagnosis. In treating these complex, extensive aneurysms, the reversed elephant trunk technique facilitates staged repair in patients who require distal aortic replacement during the first operation.
CARDIOVASCULAR
Ann Thorac Surg 2005;80:2166 –72
2172
COSELLI ET AL REVERSED ELEPHANT TRUNK TECHNIQUE
Ann Thorac Surg 2005;80:2166 –72
References
CARDIOVASCULAR
1. Crawford ES, Coselli JS, Svensson LG, Safi HJ, Hess KR. Diffuse aneurysmal disease (chronic aortic dissection, Marfan, and mega aorta syndromes) and multiple aneurysm: treatment by subtotal and total aortic replacement emphasizing the elephant trunk operation. Ann Surg 1990; 211:521–37. 2. Rokkas CK, Kouchoukos NT. Single-stage extensive replacement of the thoracic aorta: the arch-first technique. J Thorac Cardiovasc Surg 1999;117:99 –105. 3. Minale C, Splittgerber FH, Reifschneider HJ. Replacement of the entire thoracic aorta in a single stage. Ann Thorac Surg 1994;57:850 –5. 4. Massimo CG, Perna AM, Cruz Quadron EA, Artounian RV. Extended and total simultaneous aortic replacement: latest technical modifications and improved results with thirtyfour patients. J Card Surg 1997;2:261–9. 5. Borst HG, Walterbusch O, Schaps D. Extensive aortic replacement using “elephant trunk” prosthesis. J Thorac Cardiovasc Surg 1983;31:37– 40. 6. Heinemann MK, Buehner B, Jurmann MJ, Borst HG. Use of the “elephant trunk technique” in aortic surgery. Ann Thorac Surg 1995;60:20 –7. 7. Carrel T, Althaus U. Extension of the “elephant trunk” technique in complex aortic pathology: the bidirectional option. Ann Thorac Surg 1997;63:1755– 8. 8. Coselli JS, Moreno P. Descending and thoracoabdominal aneurysms. In: Cohn LH, Edmunds LH Jr, eds. Cardiac
9. 10. 11. 12. 13. 14.
15. 16.
17.
Surgery in the Adult. 2nd ed. New York: McGraw Hill, 2003:1169 – 89. Coselli JS, Conklin LD, LeMaire SA. Thoracoabdominal aortic aneurysm repair: review and update of current strategies. Ann Thorac Surg 2002;74:S1881– 4. Coselli JS, LeMaire SA, Köksoy C. Thoracic aortic anastomoses. Operative Tech Thorac Cardiovasc Surg 2000;5:259 –75. Coselli JS, LeMaire SA, Walkes JC. Surgery for acute proximal aortic dissection. Operative Tech Thorac Cardiovasc Surg 1999;4:13–32. Coselli JS, LeMaire SA. Experience with retrograde cerebral perfusion during proximal aortic surgery in 290 patients. J Card Surg 1997;12(Suppl):322–5. Coselli JS, Oberwalder PJ. Successful repair of mega aorta using reversed elephant trunk procedure. J Vasc Surg 1998; 27:183– 8. Shiiya N, Yasuda K, Murashita T, et al. Application of the reversed elephant trunk technique during total thoracoabdominal replacement in patients with Marfan’s syndrome. J Card Surg 1998;13:56 –9. Zanetti PP. Replacement of the entire thoracic aorta according to the reversed Elephant Trunk technique. J Cardiovasc Surg 2001;42:397– 402. Carrel T, Berdat P, Kipfer B, Eckstein F, Schmidli J. The reversed and bidirectional elephant trunk technique in the treatment of complex aortic aneurysms. J Thorac Cardiovasc Surg 2001;122:587–91. Crawford ES, DeNatale RW. Thoracoabdominal aortic aneurysm: observations regarding the natural course of the disease. J Vasc Surg 1986;3:578 – 82.
DISCUSSION DR JOSEPH BAVARIA (Philadelphia, PA): Doctor Coselli, that was a great paper and, like usual with your work, it shows us how we can make the total arch operation so much easier without having to construct the distal anastomosis in a patient who has already had the distal arch and descending aorta replaced. I have two questions. The first one is: what is the average time between the two operations in your experience? But the more important second question is: we are usually taught in vascular surgery, or classic aortic surgery, that you always operate proximally first and distally second, and so we usually do it that way. When do you decide to perform the distal operation first with regard to the proximal operation? Is there a size or diameter of the aneurysm proximally that you use to make the decision to operate distally first? Is there a ratio of the two aneurysm sizes or any novel kind of decision-making process that you can enlighten us with regarding whether you reconstruct proximal first or proximal second? Thank you. DR COSELLI: Thank you, Joe. This report involves a select group of patients, whom we identified for this particular approach owing to the fact that they had known pathology involving the ascending aorta or transverse aortic arch, but who needed their descending thoracic or thoracoabdominal aortas managed first on the basis of symptomatology or disproportionate size. As a consequence, almost half of the patients presented with back pain and there were several others who suffered from contained rupture. In addition, there were a group of patients in whom the ascending aorta or arch was 5 to 6 cm, but in whom the descending thoracic component was 8 to 9 cm, and were treated using the reversed elephant trunk staged fashion based primar-
ily upon the size differential as the dominant factor. The time between the two procedures was an average of 4 months. DR THOMAS WOZNIAK (Indianapolis, IN): I have a technical question. How do you cannulate these patients at the time of their second operation? Do you have concerns about “windsocking” the elephant trunk with femoral cannulation to cover the arch vessels, or do you cannulate the subclavian artery to avoid this potential complication? DR COSELLI: Our number one preference is to use the right subclavian or axillary artery. On occasion when we don’t believe that is appropriate and in patients who do not have aortic dissection, we will directly cannulate the ascending aortic aneurysm, remove the cannula at the initiation of circulatory arrest, and then use a graft with a prefabricated side branch graft (8 mm) to reinstitute antegrade perfusion. Some of these patients had previously undergone proximal aortic replacement. As I showed in my presentation, I do not believe that it is inappropriate to cannulate these previously placed ascending aortic grafts directly to establish cardiopulmonary bypass for cooling before replacing or reconstructing the aortic arch. We have avoided femoral artery cannulation precisely out of concern for windsocking the elephant trunk portion within the proximal descending thoracic aortic graft. Since we have not used the femoral approach, I cannot confirm or deny that there is real risk of windsocking. DR THORALF SUNDT (Rochester, MN): How long of a trunk do you leave? DR COSELLI: Approximately 8 to 10 cm.