An Alternative Approach Using Long Elephant Trunk for Extensive Aortic Aneurysm: Elephant Trunk Anastomosis at the Base of the Innominate Artery
Background Although a staged elephant trunk procedure has been widely used, the early mortality of the first stage operation as well as the interval mortality between operations remains unsatisfactory. We developed an alternative elephant trunk procedure to reduce mortality and morbidity.
Methods and Results Ascending aorta and arch vessels were minimally dissected. During systemic cooling, a four-branched arch graft with a sewing “collar” and a long “elephant trunk” was prepared. The ascending aorta was opened under selective brain perfusion with moderate hypothermia (25°C), and the elephant trunk was then pulled down into the descending aorta using the catching catheter introduced via a femoral artery. The elephant trunk anastomosis using the collar was made at the base of the innominate artery. The arch vessels were divided and closed at aortic stump, and grafted separately as a consequence of the very proximal site for the elephant trunk anastomosis. Between October 1998 and September 2001, 17 patients, ranging in age from 25 to 79 years (mean 67 years) with extensive aortic aneurysm underwent this operation. Preoperative cardiac complications included coronary artery disease in 5, aortic regurgitation in 3, and 3 of these 8 patients had poor left ventricular function with an ejection fraction less than 40%. Nine patients underwent a second stage operation, in 1 of them the permanent elephant trunk procedure was initially attempted but the second stage procedure was done because of increasing endo-leakage. The mean interval between operations was 8 days (range 1 to 14 days) in the remaining 8 patients. In 5 of 6 patients who underwent the permanent elephant trunk procedure, a decrease in the size of the aneurysm based on thromboexclusion was observed using serial computed tomography scans. A single stage repair was performed in 1 patient. The 30-day survival rate of all operations was 100%, however, there was 1 in-hospital death (6%) after the second operation. There was no stroke, however, paraplegia occurred after the first operation in 1 patient (6%) of the in-hospital death. No new phrenic or recurrent laryngeal nerve palsy occurred as a result of surgery.
Conclusions The present technique using a modification of the elephant trunk technique for extensive aortic aneurysm provides acceptable mortality and morbidity. The present strategy would be an alternative for the standard elephant trunk procedure in some high-risk patients with advanced age and comorbidities.
The original “elephant trunk” technique was developed by Borst and coworkers in 1983.1 Since then, a staged elephant trunk procedure has been widely used as the preferable procedure for extensive ascending, arch, and descending or thoracoabdominal aneurysms.2,3 This technique has been further modified by several groups. An early mortality rate for the first stage operation of 6% to 8% and an interval mortality of 5% to 10% have been reported.4–5 Furthermore, the potential risk of rupture of the descending aorta during the interval was emphasized.5 The great tension on the elephant trunk anastomosis at the proximal descending aorta is most likely related to the ruptures. An increase in the risk of the rupture was marked when the distal aortic arch exceeded 5 to 6 cm. Recently, Svensson and coworkers made further modification of the elephant trunk procedure to reduce the risk of a rupture after the first operation. Their modification included the elephant trunk anastomosis being placed in the aortic arch between the left common carotid and left subclavian artery.6
We previously reported a modification of the elephant trunk procedure in that the elephant trunk anastomosis was placed in the ascending aorta, at the base of the innominate artery.7 That technique was aimed to reduce the risk of a rupture in the proximal descending aorta similar to Svensson and coworkers’ concept and to ensure a fast, easy, and watertight elephant trunk anastomosis with an open distal method. Furthermore, this technique could engage the rapid second stage operation if necessary. We performed this procedure in 17 patients and analyzed the results.
Between January 1997 and September 2001, 27 patients underwent total aortic arch repair. Conventional aortic arch replacement with or without the elephant trunk technique was performed in 10 patients. In the present study, 17 patients with extensive aortic aneurysm in whom an elephant trunk anastomosis was placed in the ascending aorta at the base of the innominate artery were investigated. Indications for this technique included an aortic pathology with an extensive aneurysm that required the elephant trunk procedure for repair as well as the proximal descending aorta exceeding 5 cm. We obtained approval for the present study from the institutional review committee of Labor Welfare Corporation Osaka Rosai Hospital and informed consent from each patient.
The baseline clinical characteristics of the study population are presented in Table 1. Twelve patients were male and 5 were female, ranging in age from 25 to 79 years (mean 67 years). One had a combined pathology of atherosclerosis and dissection (Figure 1, top). Preoperative cardiac complications included coronary artery disease in 5, aortic regurgitation in 3, and 3 of these 8 patients had severe left ventricular remodeling with an ejection fraction less than 40% because of valvular cardiomyopathy in 2 patients (Marfan syndrome) and ischemic cardiomyopathy in 1 patient. Brain complications included hemiplegia in 2, lacunar infarction in 2, and cerebellar infarction in 1. One patient had chronic renal insufficiency after a previous uninephrectomy. Two patients with Marfan syndrome had a chronic aortic dissection and a degenerative mega-aorta, respectively. ⇓
Surgical Procedures (First Stage)
The surgical technique of the elephant trunk anastomosis at the base of the innominate artery was described previously (Figure 2). 7 Briefly, via a median sternotomy the ascending aorta and arch vessels were minimally dissected. Patients were placed on cardiopulmonary bypass with the arterial return to the right subclavian artery and venous drainage through the bicaval cannulas, and cooled to 25°C. While cooling the patient, a four-branched arch graft (Hemashield Gold woven double velour, Meadox) with a sewing “collar” and an “elephant trunk” as long as 15 to 20 cm in a collapsed length was prepared. The 22- to 28-mm arch graft was used according to the size of the distal descending aorta measured by the preoperative contrast-enhanced computed tomography (CT) scans. In cases of aortic root replacement, the arch graft was divided into 2 grafts such as the proximal valved conduit with 4 side branches and the distal elephant trunk. Selective brain perfusion was then instituted with an additional cannula in the left common carotid artery. The heart arrest was induced by antegrade blood cardioplegia and subsequently maintained by retrograde blood cardioplegia. The ascending aorta was transected at the base of the innominate artery. The elephant trunk was inserted into the descending aorta with a catching catheter (Amplatz Goose-Neck Snare, Microvena), which was introduced through a long sheath inserted into the left femoral artery by Seldinger technique. The anastomosis between the collar of the graft and the distal aorta was carried out at the base of the innominate artery under the open distal method. In most patients, the base of the innominate artery was less dilated and less diseased segment, so that the elephant trunk anastomosis was performed with a lack of difficulty. Approximately 15 to 20 cm of the elephant trunk was left in the distal aorta so that the distal end of the elephant trunk was located at the level of Th6 to Th8. After the distal anastomosis was accomplished, the perfusion to the lower body was resumed using the side branch of the graft. The proximal anastomosis was carried out at the distal to sinotubular ridge. During the rewarming period, the arch vessels were divided and closed at aortic stump, and grafted separately as a consequence of the very proximal site for the elephant trunk anastomosis The branched vascular prosthesis was entirely covered with expanded polytetrafluoroethylene membrane.
Surgical Procedures (Second Stage)
At the second stage operation the descending aorta was approached through a left thoracotomy. Normothermic closed femoro-femoral veno-arterial bypass with a heparin-coated circuit and oxygenator was used for proximal unloading and distal protection. The elephant trunk lying in the descending aorta was confirmed with intraoperative periaortic echography. After minimal dissection of the descending aorta the distal aortic clamp was firstly placed on an uninvolved segment of the descending aorta. A proximal aortic clamp was then placed over the aortic segment including the elephant trunk. These procedures could reduce intraoperative bleeding and facilitate the distal anastomosis. After the aortotomy, the elephant trunk prosthesis was pulled out and anastomosed to an uninvolved segment of the descending aorta. The graft was completely wrapped with the aneurysmal wall so as to reduce postoperative bleeding.
Single Stage Operation (Pull-Through Technique)
In a patient with particularly large extensive aneurysm, the single-stage operation was performed. The ascending aorta and aortic arch was replaced using the technique described above through the median sternotomy, and the descending aorta was replaced using the long elephant trunk that was pulled out through antero-lateral thoracotomy.
Permanent Elephant Trunk
Provided that the contact zone expected between the elephant trunk prosthesis and the aortic wall at the Th7 to Th8 level was approximately 3 cm or more in length, the first stage procedure was attempted as the “permanent elephant trunk” that did not require subsequent distal anastomosis in the descending aorta. Candidates for this type of operation include, for example, chronic dissection with the entry being closed by the contact zone between the graft and true lumen, a large saccular aneurysm that terminates at the proximal thoracic aorta, and recurrent proximal dissection after the graft replacement of the descending aorta. Postoperative contrast-enhanced CT scans were obtained at 1 to 3 days after the first stage operation in these patients, and a decision was made whether a second stage operation was necessary.
Eight patients underwent the rapid two-stage operation, and 5 patients underwent the permanent elephant trunk procedure. Another 1 was attempted to undergo the permanent elephant trunk procedure, but required the second stage procedure on the 34th postoperative day because of increasing endo-leakage at the distal site of the elephant trunk. For the 8 patients who underwent the rapid two-stage operation, the mean interval between the first- and second-operation was 8 days (range 1 to 14 days). In 5 of 6 patients who consequently underwent the permanent elephant trunk procedure, the obliteration of the distal aneurysmal lumen by thrombus formation along the elephant trunk was confirmed by the CT scans performed several days after the operation. Furthermore a reduction in the size of the aneurysm because of thromboexclusion was observed by the serial CT scans with a mean follow-up period of 14 months (range 3 to 28 months). One patient, a 67-year-old female, underwent urgent single stage repair. She presented with back pain, progressive dyspnea and dysphagia for solid food. CT scans revealed an extensive thoracic aortic aneurysm. The ascending aorta was 6.5 cm in diameter and the descending aorta was 12 cm where the aneurysm had eroded the vertebral column and compressed the left main bronchus (Figure 1, bottom).
Simultaneous procedures included coronary artery bypass grafting in 2, aortic valve plasty in 1, and modified Bentall’s procedure in 2. Two patients with Marfan syndrome are currently awaiting the further thoracoabdominal aortic repair.
There was no operative death within 30 days after surgery. However, there was 1 in-hospital death (6%) after the second operation: a 79-year-old patient with chronic renal insufficiency died from septic multiple organ failure because of purulent cholecystitis. There was neither stroke nor temporary neurologic deficits in the entire patients, however, paraplegia occurred after the first operation in 1 patient; the in-hospital death. No patient required re-exploration for postoperative bleeding. No perioperative myocardial infarction occurred. Twelve of 17 patients (71%) were extubated within 24 hours of the first operation. The cardiopulmonary bypass findings of the first operation are summarized in Table 3. The mean cardiac arrest time was 68 minute (range, 36 to 87 minute) in the 12 patients who did not require additional cardiac procedures. The mean open distal time and mean brain perfusion time were 29 minute (range 14 to 51 minute) and 125 minute (range 67 to 226 minute), respectively. The mean operative time of the first stage procedure including the case underwent additional cardiac procedure was 364 minute (range 282 to 486 minute). The mean operative time of the second stage procedure was 136 minute (range 117 to 155 minute). No new phrenic or recurrent laryngeal nerve palsy occurred as a result of surgery.
Even in the recent progress in adjunct methods, surgical repair for the extensive aortic aneurysms remains unsatisfactory to be associated with considerable mortality and morbidity, especially in the elderly or patients with preoperative complications. While a single-stage repair has been sporadically reported,8,9 a staged elephant trunk procedure has been widely used as the preferable procedure for extensive ascending, arch, and descending or thoracoabdominal aneurysms.1–5 However, the relevant neurologic damage, hemorrhagic complications and myocardial damage remain important factors to affect directly the operative outcome, and some patients are exposed to the risk of rupture of the remaining distal aneurysm while awaiting the next stage operation.4,5
Safi and associates, in their series of 63 patients who were operated on using the elephant trunk technique, reported that 4 patients (6%) died early after the operation and 6 patients died subsequently during the interval period between the first operation and the second operation. Three of these 6 interval deaths in their series were because of distal aortic aneurysm rupture.5 Svensson and associates reported a 30-day survival rate of 92% in 84 patients who underwent operations on the aortic arch using the elephant trunk technique, however, there were 4 patients who died in the interval between the first operation and the second operation because of rupture of the descending aorta.4 Thus, the disadvantage of the staged elephant trunk procedure for an extensive aortic aneurysm is associated with the cumulative risk of multiple operations. To reduce the interval mortality rate, it appears to be important to prevent the rupture of the residual aortic segment after the first procedure as well as to shorten the interval between the operations. Usually, the second stage descending aortic grafting has been recommended after 4 to 8 weeks of recovery. More recently, Svensson and associates have made further modification of elephant trunk procedure to reduce the risk of a rupture after the first operation in which the elephant trunk anastomosis was placed in the aortic arch between the left common carotid and left subclavian artery.6 We concur with Svensson and colleagues, and we previously advocated the elephant trunk anastomosis to be performed in the ascending aorta at the base of the innominate artery.7 The advantage of this approach over the conventional procedure is a reduction of stress on the elephant trunk anastomosis when made at a less dilated segment. Our strategy is not only to reduce the risk of rupture in the proximal descending aorta, which is hazardous in the conventional method, but also to shorten the interval based on rapid recovery after the first operation. Importantly, we could intentionally perform the rapid two-stage repair even in elderly patients ranging from 62 to 79 years (mean 73 years) with an interval of 8 days.
Selective brain perfusion (SBP) as well as deep hypothermic circulatory arrest with or without retrograde brain perfusion is the common adjunct used for the surgery of the aortic arch.4,5,10 Although no standard perfusion strategy has a statistical superiority than the others for brain protection, we used the SBP and open distal method at 25°C in the present series. This technique provides adequate time for individual repair of the arch vessels and reduces the risk of coagulopathy as in the deep hypothermia. In 5 most recent cases, the mean selective brain perfusion time was 94 minutes (range 67 to 110 minutes). We did not experience any strokes in any patients.
The elephant trunk procedure at the base of the innominate artery using the open distal method yields a secure and rapid anastomosis, and reduces the risk of aortic tearing even in the case of the suture line that is tailored down to the smaller size of the graft. The prolonged period of the open distal procedure can potentially result in distal organ failure such as spinal cord injury and renal failure. The safe limit of the open distal method has been suggested to be 30 minutes at 25°C. In cases of our learning period, a 79-year-old male suffered from paraplegia after the first operation, in that case the prolonged open distal time of 44 minute could be a possible cause for paraplegia. He was the only patient who required hemodialysis. However, we have refined this technique so that the open distal time was 14 to 18 minute in 5 most recent cases.
Reattachment of the arch vessels as a Carrel patch has been widely used for arch repair as a simple and less time-consuming method. However, this island technique appears to have a disadvantage of the diseased aortic wall to be unresected, and the diseased aortic wall may reexpand in certain cases of aortic dissection. In addition, we often found severe atheromatous change inside the aortic arch, especially around the orifice of arch vessels. Recent studies emphasized that the atherosclerotic aortic wall is a possible source of brain infarction.11 Kazui and coworkers reported the excellent results of the total arch replacement using arch branched graft and selective brain perfusion for atherosclerotic arch aneurysms.10 For these reasons, our aortic arch repair was comprised of the complete replacement of the aortic arch and individual reconstruction of arch vessels.
The length of the ET prosthesis was advocated to be no more than approximately 15 cm. Borst and coworkers described a thromboexclusion technique using a long elephant trunk in a patient with an atheroscrelotic descending aortic aneurysm.1 This concept of the “permanent elephant trunk” was further developed in several forms. Ando and associates reported that the elephant trunk inserted into the true lumen of the descending aorta could promote closure of the false lumen inpatients with aortic dissection.12 The stented elephant trunk procedure has also been introduced.13,14 A considerable risk of paraplegia exists in the placement of a long elephant trunk. In the present study, the appropriate length and diameter of the elephant trunk was determined by preoperative aortography and computed tomography. As a result, the distal end of the elephant trunk was located at the level of Th6-Th8. An additional risk of insertion of a long elephant trunk is potential dislodging thrombus in the descending aorta and embolism of the visceral arteries or lower extremities. To avoid embolic complications, patients with the mobile thrombus in the descending aorta should be excluded from this procedure. We routinely used transesophageal echography for quantitative analysis of the thoracic aorta. Furthermore, to remove all possible atheromatous plaques and thrombi we performed the retrograde perfusion from the femoral cannula at a flow rate of 1.5 L/min for 1 minute after the elephant trunk anastomosis.
The Limitations of the Present Study
There are indeed some disadvantages associated with this procedure. Encircling arch vessels as well as cannulation for selective brain perfusion have all potential risks for stroke. Insertion of the long elephant trunk has certain risks of paraplegia and distal embolism. Thromboexclusion; even in the sutured elephant trunk, could not be promised, and the residual intercostal arteries may be possible factors to cause reexpansion of the descending aorta. In addition, this procedure leaves the patient with branched vascular prosthesis behind the sternum. Therefore, the risk of infection may increase, and resternotomy itself will be troublesome in case of reoperation through median sternotomy. This procedure will be open to further modifications including operative indications and techniques. Our favorable experience was obtained using a relatively small numbers of patients, and has to be confirmed with a larger series of patients.
The present approach using a modification of the elephant trunk technique for extensive aortic aneurysm would be a potential alternative for the standard Borst technique. It is suggested that there are some patient categories where this strategy will provide less mortality and morbidity.