Endovascular stent-grafting via the aortic arch for distal aortic arch aneurysm: an alternative to endovascular stent-grafting

Abstract

Objective: We have experienced transaortic stent-grafting for treating distal arch aneurysm or type B dissection. This paper is to mainly report the surgical aspect of these procedures. Methods: Fifteen patients underwent this surgery, including 12 men and three women ranging from 47 to 83 years. Twelve had aneurysms and three aortic dissection. Concomitant surgery was necessary in seven patients (coronary artery bypass grafting in five, tricuspid annuloplasty in one, and replacement of ascending aorta and/or total arch replacement in three cases). A stent graft (Gianturco Z-stent and Intervascular prosthesis) was loaded in a 30-F sheath catheter. Under circulatory arrest, selective cerebral perfusion was established, and the sheath catheter was inserted through aortotomy into descending aorta and the stent graft was deployed at an appropriate level. The proximal end of graft was sutured to the aorta just distal to the left subclavian artery with inclusion method at the posterior wall. Concomitant surgery was done during cooling or rewarming period. TEE was utilized to visualize every endovascular manipulation to avoid unintended intimal injury or misplacement of graft and to assess the surgical results in the operative theater. Results: Aneurysm was successfully excluded except in one patient who had a proximal endoleak and distal endoleak due to underestimation of aortic diameter. There was one operative mortality caused by cerebral infarction, possibly due to debris from femoral arterial cannulation. In the remaining patients, there was no enlargement of residual aneurysm. The excluded aneurysmal sac gradually regressed and disappeared within 2 years in five patients and the thrombosed false lumen completely shrunk within 1 year in two patients. One patient had paraplegia, possibly because the graft was intentionally advanced deeply to cover the thick and fragile atheromatous layer in order to avoid destruction of the atheroma by an expanded graft. Conclusions: Endovascular stent graft via the aortic arch is an acceptable treatment for distal arch aneurysms close to or involving left subclavian artery or type B dissections, especially for those cases requiring other cardiac procedures. It can lead to regression and disappearance of aneurysm or dissection in the mid-term follow-up.

Aortic aneurysm, Stent grafting, Aortic dissection, Transesophageal echocardiography

1 Introduction

Three procedures are available for treating distal arch aneurysm or type B dissection: (1) left thoracotomy approach; (2) median sternotomy approach; and (3) catheter intervention. In the left thoracotomy approach with or without partial bypass, aortic cross-clamping is necessary, which can cause paraplegia or new intimal tear at the clamp site. In median sternotomy, anastomosis can be done without cross clamping, ‘open distal anastomosis’, under deep hypothermic circulatory arrest together with selective cerebral perfusion (SCP). However, deep and poor visual field makes the distal anastomosis and hemostasis around it difficult. Recently developed catheter intervention [1,2] has reduced surgical stress and risk of neurological complications during SCP, hypothermia, and retrograde femoral perfusion, and has widened its indication to elderly patients with higher risks. In this method, however, occurrence of endoleak and misplacement of graft remains problems, especially when the proximal neck is short. When the patients necessitate other surgical treatments such as coronary artery bypass grafting (CABG) or arch replacement, as is often the case in this category of patients, left thoracotomy and catheter intervention necessitates two-staged procedures.

We have developed endovascular stent grafting via the aortic arch [3], ‘open-stent grafting’, after modifying Kato's method in which a bypass from the ascending aorta to the arch branches is prepared before placing the stent graft into the descending aorta [4]. In this report, we are describing the practical aspect of this method, mainly focusing on the surgical techniques including pitfalls, advantages and limitations, and clinical mid-term results (1–3 years) of this method from our initial experience.

2 Subjects and methods

Fifteen patients underwent open stent-grafting between September 1997 and October 1999. They included 12 men and three women with ages ranging from 47 to 83 years (65.8±14.8 years) with seven patients over 70 years old. The diagnosis was aneurysm in 12 (saccular type in all) and dissection in three patients. This procedure was indicated in patients with distal aortic arch aneurysm which are situated too close to the left subclavian artery with too short a proximal neck (within 1 cm) for transluminal stent grafting, or which are associated with other pathologies necessitating surgical intervention. When the descending aorta was diffusely dilated and there was no appropriate attachment portion, this procedure was not indicated. Concomitant surgery was necessary in seven patients, including CABG in five, tricuspid annuloplasty in one, and replacement of ascending aorta and/or total arch replacement in three patients. In two patients, left subclavian artery was involved in or so close to the aneurysmal portion and was sacrificed. An 8-mm ePTFE graft was used as a bypass to this artery from the right subclavian artery. The Institutional Ethics Committee on Human Research in our institute approved this study and informed consent was obtained from every patient.

After induction of anesthesia, a 5-MHz biplane transesophageal echocardiography (TEE) was introduced and was used throughout surgery for reconfirming the preoperative diagnosis, measuring the aortic size around the attachment portion, guiding endovascular procedures, and assessing the surgical results immediately following termination of cardiopulmonary bypass (CPB) as well as being used as a routine intraoperative monitor. Associated surgical procedures were completed while the patient was cooled down to 25°C rectal temperature.

The CPB was established with an arterial cannula in the ascending aorta as long as it was feasible. When the ascending aorta was considerably diseased, based on TEE assessment, the arterial cannula was placed in the femoral artery in the early cases. However, after we experienced broad cerebral infarction in case #3, we changed the arterial access to the right subclavian artery by using an 8-mm ePTFE graft anastomosed in an end-to-side fashion.

A stent graft was prepared using a Gianturco self-expandable stainless steel Z-stent (Cook Inc., Bloomington, IN) and a vascular prosthesis (Intervascular Inc., Clearwater, FL). The graft size was determined according to the size measured with computed tomograms and TEE views. A graft of 0–4 mm larger size was selected.

The stent graft was loaded in a 30-F flexible sheath catheter. The catheter tip was mildly bent so as to follow the curvature of distal arch. Under circulatory arrest at the rectal temperature of 25°C, with SCP established, the anterior aortic wall was incised at just distal to the left subclavian artery (Fig. 1A) . In cases of aortic dissection and those with concomitant arch replacement, the aorta was transected. After the aortic wall and aneurysmal portion or entry site was inspected and proximal suture line was determined, the sheath catheter loaded with stent graft was inserted into the descending aorta while TEE visualized both the aorta and sheath catheter. Immediately before deploying the stent graft, the level of catheter tip was assessed by the distance from the level of diaphragm. After the TEE probe was advanced into the stomach, it was gradually withdrawn to the level where the diaphragm between the liver and right atrium was depicted at the midline of TEE image in longitudinal scan (level A). Then the TEE probe was further withdrawn while visualizing the descending aorta up to the level where the catheter tip in the aorta and distal end of aneurysm was depicted at midline of TEE image in longitudinal scan (levels B and C, respectively). We tried to keep the difference of depth of probe tip between levels A and B larger than 9 cm (corresponding to height of approximately 3–4 vertebrae) and that between level B and C (length of attachment portion) more than 5 cm. The stent graft was slowly pushed out of the catheter with a pushing rod by holding the pushing rod steadily and withdrawing the sheath catheter so that the anticipated level of graft would not alter. Deployment of graft and its expansion in the aorta was visualized in real time with TEE.

Fig. 1

Schematic illustration of transaortic stent-grafting procedures. (A) Aortotomy is put just distal to the left subclavian artery. A sheath catheter loaded with stent graft is inserted into descending aorta. (B) The proximal end of the graft is sutured to the aortic wall. (C) The aortotomy is closed.

The distal end of graft often appeared as curved inward after it was released out of the sheath catheter. The edge was tailored by using an inflated balloon which compressed the graft toward the aortic wall. This procedure was also navigated by TEE views to avoid an excessive and/or misplaced inflation of balloon.

The graft was lightly pulled to ensure that stented portion firmly attached to the aorta. An excessive portion of unstented graft which appeared out of aortotomy was transected and trimmed while carefully eliminating a twist of graft. The proximal end of graft was sutured to the aortic wall: the posterior half with inclusion method and the anterior half sutured to the incised edge of aorta (Fig. 1B). Perfusion to the descending aorta was resumed antegradely through a balloon catheter placed in the graft with the balloon inflated (at a flow rate of 1 or 2 l/min). The aortotomy was closed with a continuous suture of 4-0 prolene, reinforced by a pair of felt strip outside the aorta (Fig. 1C). In cases of aortic dissection and those with arch replacement, the transected edge of aortic wall was reinforced with internal graft and external felt strip, then the aorta was anastomosed to the branched graft.

As the systemic circulation was resumed and the patient was rewarmed toward the body temperature of 36°C, the blood flow in the graft and endoleak was examined with TEE. The ePTFE graft anastomosed to the right subclavian artery was utilized as a graft to the left subclavian artery when the latter artery was sacrificed.

After surgery, position of the stent was examined by chest X-ray in the intensive care unit. By means of postoperative computed tomography (CT) scan and angiography, adequacy of graft size, endoleak, leak at the proximal anastomosis, and thromboexclusion of aneurysmal lumen or false lumen was assessed.

All data are expressed as mean±standard deviation.

3 Results

The graft size ranged from 22 to 32 mm (27.5±3.2 mm) according to the diameter of attachment portion in the computed tomogram and/or TEE views. The CPB time was 162–313 min (238.2±47.7 min), the SCP time 58–180 min (104.1±42.1 min), and circulatory arrest time 55–95 min (68.1±11.6 min). The time from termination of CPB to chest closure was 50–262 min (107.2±60.2 min).

Fig. 2 shows pre- and postoperative angiograms of a successful case (case #9). Aneurysmal lumen which was situated at just distal to the left subclavian artery (Fig. 2A) was successfully excluded after surgery (Fig. 2B). Such a result was obtained except in one case (case #6). The stent portion is found distal to the aneurysmal portion and is situated at the upper half of descending thoracic aorta.

Fig. 2

Angiograms of a successful case. (A) The aneurysm is situated just distal to the left subclavian artery. (B) the stent graft has been excluded by this graft. The stent portion works as an attachment portion.

We experienced one hospital death due to broad cerebral infarction (case #3) probably caused by debris from iliac artery or by thrombus in the thoracic aneurysm, and one late death (8 months after surgery) caused by aspiration pneumonia unrelated to the surgical procedures. Case #11 had paraplegia postoperatively, possibly caused by a too deep insertion of graft (the distal end of graft situated at only 7 cm from diaphragm). In this particular case, the graft was intentionally advanced deeply to cover the thick and fragile atheromatous layer which had been found with TEE, in order to avoid destruction of the atheroma by an expanded graft although a too deep insertion had been apparent in TEE assessment before deployment.

There was one case with endoleak (case #6). Intraoperative TEE had already revealed a gap around the stented portion of graft and blood flow there (Fig. 3A) . Underestimation of aortic diameter was deemed to be responsible. TEE had also shown leakage at the proximal anastomosis site (flow into the aneurysmal sac around the graft). Although the second pump run was indicated, we hesitated using it because of a high risk of prolonged pump run in this case. Endoleak as well as proximal leakage was confirmed by postoperative CT and angiography (Fig. 3B). The proximal endoleak probably occurred at the small gap at the inclusion suture line. In other cases no apparent endoleak was found with both TEE and postoperative examination.

Fig. 3

(A) Intraoperative transesophageal echocardiogram showing a gap around the graft, indicating endoleak (case #6). (B) Postoperative angiography apparently shows proximal leakage and endoleak at the distal end.

In the mid-term follow-up, the aneurysm sac was found eventually to be regressed except in one case with endoleak (case #6). The excluded aneurysmal lumen completely disappeared in five cases. Fig. 4 shows changes of aneurysmal lumen in case #1 after 16 months of surgery. The atheromatous layer (Fig. 4A) has disappeared to leave an apparently normal aortic wall (Fig. 4B). In case #5 with aortic dissection, the false lumen was thrombosed immediately following stent grafting (Fig. 5A,B) . After 8 months, this false lumen has completely disappeared and the aortic wall appears as normal (Fig. 5C).

Fig. 4

Computed tomograms showing disappearance of aneurysm after stent grafting (case #1). (A) Preoperative; (B) 16 months after surgery.

Fig. 5

Computed tomograms showing thrombosis of false lumen (FL) following stent-grafting (A,B) and disappearance of thrombosed FL (C).

4 Discussion

Transaortic stent grafting has solved several problems of conventional methods. It avoids aortic cross-clamping which can damage the intima and possibly cause new intimal tear at the anastomosis site. It spares distal anastomosis in the deep and poor visual field, thus minimizing the bleeding around the distal anastomosis. Aortic clamping can be avoided as well if the aorta is replaced under circulatory arrest via left thoracotomy or transverse sternotomy. However, our procedures spare pleurotomy and manipulation of lung, thus minimizing the damage on the lung and an occurrence of airway bleeding. In this series there was no new development of airway bleeding. The difference from Kato's original method [3] is that no bypass is placed from the ascending aorta to the arch branches. Anastomosis with side clamping of ascending aorta can be avoided in the patient group of frequent occurrence of atheromatous changes in the entire aorta.

Problems of this method would be a number of blind maneuvers in the descending aorta, which potentially cause intimal injury at the time of inserting a sheath cannula or misplacement of catheter tip, leading to inadequate length of attachment portion or too deep insertion of graft causing paraplegia. We have solved these problems by using TEE which clearly visualizes every intraluminal event [5]. It was also useful: (1) for assessing the atheromatous changes of aortic wall around the aneurysm, spatial relationship of entry site to the branch arteries; (2) for guiding surgical procedures; and (3) for assessing the surgical results immediately following reperfusion in the operative theatre. Although the arch branches have been considered as a blind zone for TEE, we have developed a method of visualizing arch branches with intraoperative TEE in the majority of cases [6].

Although catheter intervention is expected to be applied to a larger variety of pathologies in the near future, surgical open-stent grafting is advantageous when the patient needs concomitant surgeries such as coronary revascularization, valve surgery, or other aortic replacement, which are not uncommonly associated to the distal arch aneurysm, thus enabling ‘one-stage procedure’. When this procedure is combined with total arch replacement, it may be a more distally secure alternative to the elephant trunk procedures in some patients. It should be stressed that these additional procedures does not prolong the bypass time because these procedures can be finished during the cooling or rewarming period which takes nearly half an hour.

The proximal anastomosis was done at the less diseased portion of aorta upon inspection of intimal aspect of aorta following aortotomy. However, aneurysm was occasionally situated too close to the origin of left subclavian artery. A suture placed too close to the left subclavian artery may lead to malperfusion of this artery. In such a case, we sacrificed this artery and a bypass was placed from the right subclavian artery by using the ePTFE graft which was also used for an arterial line during CPB. We did not make a bypass between the left subclavian artery and left common carotid artery because of possible occurrence of cerebral infarction during this procedure.

Interestingly, the residual aneurysmal lumen or false lumen which was excluded by this technique was found to shrink and finally disappear within 2 years. It seems that exclusion has eliminated pressure loading to the residual lumen and the subsequent hemostatic and healing process has gradually corrected morphology of the aorta within a few years. In the case with significant endoleak, there was no reduction of aorta size until he died of aspiration pneumonia 8 months postoperatively.

The most significant complication of this method was cerebral infarction. There are several possible etiologies for this event: debris flushed by the jet stream from the arterial cannula placed in the femoral artery or mural thrombus at the aneurysmal wall; malperfusion of arch branches during SCP; air embolism; and so on. The first one can be eliminated as the arterial cannula placed in the right subclavian artery to generate antegrade blood flow in the aorta. The second one can be minimal by careful manipulation during dissecting the aorta. Paraplegia is another complication. In this patient, the graft position was too deep and perfusion pressure during SCP was as low as around 30 mmHg. This situation may be prevented by elevating the perfusion pressure and placing the graft at rather high level. Transaortic stent grafting does not yet solve the disadvantage that it necessitates circulatory arrest and manipulation/cannulation of arch vessels. Possible complication of this procedure involves injury on the esophagus due to extensive manipulation of TEE probe. Fortunately, we experienced no complication related to TEE manipulation such as gastrointestinal bleeding in spite of extensive use of it. No patient complained of gastrointestinal tract disorders which necessitated endoscopic examination postoperatively. We consider that an experienced TEE operator can safely perform this procedure. However, further investigation is mandatory to improve the clinical outcome without any complication.

In conclusion, endovascular stent graft via the aortic arch is an acceptable alternative treatment for distal arch aneurysms or type B dissections, especially for those cases requiring other cardiac procedures. It can lead to regression and disappearance of aneurysm or dissection in the mid-term follow-up.

Appendix A Conference discussion

Dr M. Sarsam (Belfast, UK): I have one question and one comment. Two years ago we described a technique which is almost similar to yours but without the use of a stent in which the aorta is opened, as you opened it, just distal to the arch, and then also posterior to the pericardium, and then a graft is put in surgically, stitched both distally and proximally, which will avoid the problem of trying to insert a stent, because I noted that your circulatory arrest time, cerebral perfusion time is rather long. The question actually is what is the length of the stent you used in your cases?

Dr Orihashi: Regarding your comment, we haven't experienced your technique. As you pointed out, the circulatory arrest time is long. But as we learn more, it will become shorter. Because our technique solves the problems of other techniques, we are thinking of continuing this procedure. Regarding your question, the length of each stent is 2.5 cm. We use double or triple stents to make the attachment portion of 5–7 cm. By using TEE guidance, the stent graft can be inserted into the aorta easily and safely.

Dr R. Griepp (New York, NY, USA): Could you tell us if the patient you described was symptomatic, that is was he having any pain arising from the aneurysm? The reason I ask is that this is a relatively small localized aneurysm that in the absence of symptoms would be unlikely to rupture and which in our practice we would follow with periodic CT scans rather than undertake surgery. I think we are seeing an increasing number of patients who undergo stent grafting in part because they are technically amenable to this technique. Our studies of the natural history of aneurysms in the descending aorta suggest that a small localized asymptomatic aneurysm has a quite low probability for rupture.

Dr Orihashi: Do you mean chest pain?

Dr Griepp: Yes.

Dr Orihashi: No case in our series had chest pain at the time of surgery except in the dissection cases.

Dr Griepp: I am asking if the patient had experienced any pain because a localized aneurysm 4–4.5 cm in diameter that is not symptomatic is unlikely to rupture and therefore there is no indication for operation.

Dr Orihashi: Sorry, I don't really understand.

Dr Turina (Zürich, Switzerland): The question is, why did you do anything to this patient?

Dr Orihashi: Why did I?

Dr Turina: He didn't need any treatment. That is the point which Dr. Griepp is making.

Dr Orihashi: You mean conservative therapies?

Dr Turina: Yes.

Dr Orihashi: Like catheter intervention? Right now the radiologist in our hospital is not so active regarding catheter intervention for thoracic aortic aneurysm. That is the main reason.

Dr J. Bachet (Paris, France): I might be a little sleepy this morning, but my point is that I don't get exactly the concept of your operation. I observe that it is rather invasive, you open the chest, you go on bypass, and you go on deep hypothermia. So why don't you simply replace the arch?

Dr Orihashi: Several patients were indicated for surgery because of associated cardiac diseases to be treated surgically. Rather than two-staged operations, we chose a combined operation, because the associated surgery could be finished during the cooling period. The aneurysm size was over 5 cm in every case without concomitant surgery. They were mostly over 75 years of age. Even if we can follow up the growth of aneurysm for years, the surgical risk steadily increases.