Long-Term Results of Open Stent-Grafting Using a Matsui-Kitamura Stent to Treat Thoracic Aortic Aneurysm

Abstract

Purpose: We describe a retrospective study of initial and long-term outcomes with an open stent grafting (OSG) with a Matsui-Kitamura stent for treating thoracic aortic aneurysm.

Methods: Between August 2005 and September 2013, 50 patients with aortic arch disease extending to the descending aorta underwent OSG. Circulatory arrest with total cardiopulmonary bypass and selective cerebral perfusion were used, and the aorta was transected between the brachiocephalic and left subclavian artery. The stent-graft was inserted, sutured to a transected aortic edge, and anastomosed to a four-branched arch graft. Preoperative, operative, and short- and long-term postoperative data were obtained from the patients’ medical records.

Results: The perioperative (within 30 days) mortality rate was 8%. Two patients (4%) had a stroke and 5 patients (10%) had a spinal cord injury resulting in paraplegia or paraparesis (1 patient each) or transient paraplegia (3 patients). Actuarial survival rates at 1, 3, 5, and 7 years postoperatively were 87.8%, 78.3%, 70.7%, and 65.3%, respectively; the rates of freedom from an aortic event were 100%, 89.1%, 82.2%, and 74.7%. There were no complications related to use of the stent-graft.

Conclusion: Our OSG method provided durable results in patients treated for thoracic aortic aneurysm, with few adverse events.

Introduction

Recent advances in the surgical treatment of thoracic aortic aneurysm (TAA) including arch aneurysm have resulted in acceptable outcomes, but patients with extensive disease and comorbid conditions continue to have relatively high mortality and morbidity rates.^1,2) The open stent-grafting (OSG) technique,³⁾ which has been described as the frozen elephant trunk or stented elephant trunk technique in Europe and America, can be performed through a median sternotomy without a left thoracotomy. In OSG, the distal anastomosis is achieved by means of the radial force exerted by the stent and the proximal anastomosis by suturing, in cases in which stent anastomosis would be problematic. The likelihood of phrenic or recurrent laryngeal nerve injury is thereby decreased, and the risk of postoperative pulmonary complications may also be reduced.^4–7)

However, the basic frames of most devices used in OSG have a Z configuration, which may not provide maximal device flexibility, ease of deployment, or avoidance of device migration in the long term. To address this problem, we developed an OSG method that employs a Matsui-Kitamura (MK) stent to fully support the graft.^8–10) We report a retrospective study of the initial efficacy and long-term results of this hybrid procedure in the treatment of TAAs involving the aortic arch and descending aorta.

Patients and Methods

Patients

Between August 2005 and September 2013, 50 patients with aortic arch disease extending to the descending aorta or with aortic dissection requiring aortic arch replacement were treated with the OSG technique in our unit. Our hospital’s ethics committee approved use of the operation, and informed consent was obtained from each patient after the possible advantages and disadvantages of the procedure in relation to conventional graft replacement had been explained.

The indications for OSG were an aneurysm that would have been difficult to treat by using a distal anastomosis through a median sternotomy and various characteristics that would put the patient at high risk of complications from conventional surgery: age of 75 years or older; concomitant presence of cerebrovascular disease, ischemic heart disease, chronic obstructive pulmonary disease, or renal dysfunction; and a history of thoracotomy. Table 1 shows demographic, aortic disease, comorbid, and previous vascular surgery data for the 50 patients who underwent OSG.

Table 1.

Characteristics of 50 patients with extended thoracic aortic aneurysm who underwent open stent-grafting using a Matsui-Kitamura stent

Variable	Valuea
Mean age, y (range)	70 (45–86)
Sex: male/female	37 (74)/13 (26)
Aortic arch disease
Degenerative (no rupture)	35 (70)
Degenerative (with rupture)	3 (6)
Aortic dissection: type A/type B	6 (12)/4 (8)b
Degenerative plus type B dissection	2 (4)
Mean diameter of aneurysm, mm (range)	66.5 (45–100)
Preoperative comorbid disease
Stroke	9 (18)
Ischemic heart disease	5 (10)
Chronic obstructive pulmonary disease	10 (20)
Renal dysfunction (creatinine >2 mg/dL)	6 (12)
Previous vascular surgery
Thoracic endovascular aortic repair	3 (6)
Endovascular aortic repair	3 (6)
Open thoracoabdominal aortic aneurysm repair	1 (2)
Open abdominal aortic aneurysm repair	5 (10)
Bypass surgery for peripheral artery disease	2 (4)

^aData are shown as number (%) unless otherwise indicated.

^bTwo patients had an acute and four a chronic type A dissection; all four type B dissections were chronic.

Stent-graft device

The stent-graft is constructed by a curved MK stent and a noncoated polyester fabric graft. Each device was custom made. The diameter and length of the stent-graft were determined preoperatively with use of multislice computed tomography (CT) or diagnostic imaging employing a catheter with a marker. The graft diameter was oversized 110% to 120% in relation to the diameter of the nondiseased segment of aorta below the aneurysm; the graft length was equal to the distance from the transection site to the nondiseased segment of the descending aorta below the aneurysm. The MK stent was oversized 110% to 120% in relation to the stent-graft diameter. The structural characteristics of an MK stent are such that its effective diameter is smaller than its nominal diameter when the device is extended to a length corresponding to the expected length of the graft.

Surgical procedure

The OSG method using an MK stent has been described previously.¹⁰⁾ In brief, both axillary arteries are exposed through a median sternotomy. An 8-mm graft is anastomosed to the right axillary artery, and an arterial perfusion device is inserted into the graft. A venous cannula is inserted into the right atrium, cardiopulmonary bypass is initiated, and the body temperature is cooled to 25°C (bladder temperature). During cooling, an 8-mm ringed graft is anastomosed to the left axillary artery. After ventricular fibrillation occurs, a vent tube is inserted from the right superior pulmonary vein into the left atrium. The origins of the left common carotid and axillary arteries are ligated, and selective cerebral perfusion is performed through the right axillary and left common carotid and axillary arteries (flow rate, 1 mL/kg/min). Total cardiopulmonary bypass with selective cerebral perfusion is then established (at 25°C). Intermittent antegrade or retrograde cold-blood cardioplegia is used to provide myocardial protection.

The procedure is done with the patient under circulatory arrest. The insertion depth (distance from the transection site to distal placement site) is determined before surgery with use of multislice CT or angiography, and the stent-graft is placed according to a mark made on the device at that time. In the last 29 patients in our series, with the aim of preventing ischemic spinal injury and protecting major abdominal organs during lower-body circulatory arrest, distal retrograde perfusion was established from the femoral artery by occluding the stent-graft with a balloon catheter (Equalizer; Boston Scientific, Clonmel, Ireland). Distal retrograde perfusion was not used in the first 21 patients; however, if the stent-graft was intended to be deployed at the Th9 level or lower, cerebrospinal fluid drainage was performed the day before surgery.

After the stent-graft is inserted, the aortic stump is anastomosed to an arch graft with four branches. Rewarming is started after systemic perfusion from a branch of the graft has been established. Subsequently, the proximal segment of the four-branched graft is anastomosed to the stump of the ascending aorta. The brachiocephalic artery, left common carotid artery, and the graft that was anastomosed to the left axillary artery (and introduced into the mediastinal space through the left second intercostal space) are each anastomosed to a branch of the graft. Selective cerebral perfusion and extracorporeal circulation are then discontinued.

Follow-up and data analysis

Follow-up assessments using contrast CT imaging were performed before hospital discharge, 6 months after surgery, and yearly thereafter. Major adverse events that occurred perioperatively (within 30 days of surgery) and during follow-up were recorded. Continuous data were expressed as mean ± standard deviation and categorical data as numbers and percentages. Patient survival and freedom from an aortic event during follow-up were assessed by using Kaplan-Meier analysis.

Results

Operative data and early outcomes

Table 2 shows operative data, including the mean operating, bypass, clamping, perfusion, and arrest times; concomitant procedures; and stent-graft landing zones for the 50 patients. The stent-graft was delivered successfully in all patients (technical success rate, 100%), and complete thrombosis of the aneruysm or entry closure with stable thrombosis in the false lumen were observed in 47 (clinical success rate, 94%). Adverse events that occurred within 30 days after surgery are listed in Table 3. Four patients died during their operation or in the hospital after their procedure. In the one patient who died in the hospital, the landing zone for the stent-graft was at the Th12 level. Paraplegia developed perioperatively, and the patient died from a sudden brainstem infarction 35 days after surgery. All three patients with transient paraparesis recovered sufficiently to be able to walk unaided. There were no complications related to use of the stent-graft.

Table 2.

Operative data for 50 patients with extended thoracic aortic aneurysm who underwent open stent-grafting using a Matsui-Kitamura stent

Variable	Valuea
Elective/emergency surgery	44 (88)/6 (12)
Operating time, min	364 ± 66
Cardiopulmonary bypass time, min	179 ± 21
Aortic clamping time, min	86 ± 21
Selective cerebral perfusion time, min	103 ± 23
Lower-body circulatory arrest time, min
Without retrograde distal perfusion (n = 21)	47 ± 10
With retrograde distal perfusion (n = 29)	15 ± 3
Preoperative cerebrospinal fluid drainage	3 (6)
Concomitant procedure	4 (8)
Mitral valve replacement	1 (2)
Coronary artery bypass grafting	1 (2)
Implantation of pacemaker	2 (4)
Distal stent-graft landing zone level
Th6	4 (8)
Th7	15 (30)
Th8	24 (48)
Th9	6 (12)
Th12	1 (2)

^aContinuous data are shown as mean ± standard deviation and categorical data as number (%).

Table 3.

Adverse events within 30 days in 50 patients with extended thoracic aortic aneurysm who underwent open stent-grafting using a Matsui-Kitamura stent

Event	No. (%)
Death during procedure	3 (6)a
In-hospital death	1 (2)b
Stroke	2 (4)
Spinal cord injury	5 (10)
Paraplegia/paraparesis/transient paraparesis	1 (2)/1 (2)/3 (6)
Mediastinitis	1 (2)
Re-exploration because of bleeding	1 (2)
Respiratory failure (intubation time >72 hr)c	12 (24)
Renal failure (requiring temporary hemodialysis)	1 (2)
Ischemic colitis	1 (2)

^aThe causes of death were multiple-organ failure (1 patient) or ischemic colitis (1 patient) after aneurysm rupture and respiratory failure after aneurysmal degeneration (1 patient).

^bPatient died of stroke (brainstem infarction).

^cTwo tracheotomies were required.

Late outcomes

Seven patients died during the observation period, of various causes (Table 4). One patient died after rupture of a treated chronic type B aortic dissection that was caused by a dilated false lumen from the re-entry flow (treatment failure). On Kaplan-Meier analysis, the actuarial patient survival rates at 1, 3, 5, and 7 years postoperatively were 87.8%, 78.3%, 70.7%, and 65.3%, respectively (Fig. 1). Five patients had an aortic event (new TAA or TAA rupture) during follow-up (Table 4). The rates of freedom from an aortic event at 1, 3, 5, and 7 years were 100%, 89.1%, 82.2%, and 74.7% (Fig. 1). No complications related to the OSG procedure, including migration or endoleak, occurred during follow-up (Table 4).

Table 4.

Adverse events during long-term follow-up in 50 patients with extended thoracic aortic aneurysm who underwent open stent-grafting using a Matsui-Kitamura stent

Event	No. (%)
Death	7 (14)
Rupture of treated thoracic aortic aneurysm	1 (2)
Stroke	1 (2)
Respiratory failure	1 (2)
Pneumonia	1 (2)
Esophageal cancer	1 (2)
Intestinal bleeding	1 (2)
Renal failure	1 (2)
Aortic event	5 (10)
Rupture of treated thoracic aortic aneurysm (treatment failure)	1 (2)
Rupture of other thoracic aortic aneurysm	1 (2)
New descending thoracic aortic aneurysm (not treated site)	3 (6)
Additional aortic procedure (thoracic endovascular aortic repair)	4 (8)a

^aFor a ruptured (1 patient) or nonruptured (3 patients) other or new descending thoracic aortic aneurysm.

Fig. 1.

Overall survival and freedom from aortic events of all 50 patients.

Follow-up CT images were available for 36 patients (72% of patients in the series) who survived for longer than 12 months. Analyses of these images showed that the aneurysm increased in size (>5 mm) in 1 patient (2.8%), remained unchanged (–5 mm to 5 mm) in six patients (18.7%), and showed skrinkage (–5 mm or more) in 29 (80.5%).

Discussion

Most devices used in OSG allow correct placement, with a narrow surgical view, of a large-diameter graft with a basic Z-stent design. However, new devices and methods that will help to improve outcomes of distant grafting are needed. Because currently available Z stents are inflexible, in extended TAA surgery, two or three such stents (50–75 mm) are inserted at the end of the graft to follow the aortic curve; other segments of the graft are not supported by the stent.^3–7) Complications may arise if the graft becomes twisted at insertion, a flexed stent is placed in a curve, or the terminal end of the stent migrates because of changes in the aneurysm over time.⁶⁾

The MK stent used in our series is flexible and easy to place, and it extends along and supports the full length of the graft.¹⁰⁾ MK stents are made from a single, 0.3- to 0.4-mm-diameter, superelastic nitinol wire having a transformation temperature of less than 0°C.⁸⁾ An MK stent can be inserted into a small-diameter sheath: our delivery system uses an 18F sheath and is flexible. The risk of damage to the vessel wall during insertion may be reduced because the tip of the stent is bent and the graft is completely supported. MK stents have previously been used by others in transcatheter stent-grafting^8,9) but not, to our knowledge, in OSG.

Stents placed in the distal arch are likely to be subjected to a high external pressure. MK stents have been found to be stronger than Z stents⁸⁾ and, because of their greater flexibility, may be less likely to fracture. In our series of patients who underwent OSG, no stent-grafts showed damage during follow-up. One possible disadvantage of MK stents is that the selection of the length to be used depends on the difference in diameter between the graft and stent, and this may increase the risk of inaccurate placement. In OSG, however, only the end of the stent is fixed; thus, shortening during placement is less of a concern than with the transcatheter method. Nevertheless, the development of new methods that address the shortening problem will improve the safety and accuracy of device placement.

Spinal cord ischemia is a devastating complication of OSG. A review of six OSG series (210 patients overall) by Uchida, et al.¹¹⁾ found that the mean rate of onset of paraplegia after OSG was 8.2% (range: 2.9%–24%). In our series, paraplegia occurred in 1 patient (2% rate), an 86-year-old man in whom the stent-graft was placed at the Th12 level and mitral valve replacement was performed concurrently. Spinal cord injury during OSG has been associated with many factors, including thoracic vertebral level, the site of deployment of the distal end of the frozen elephant trunk, a history of downstream aortic surgery, intraoperative hypotension, and the type of aortic disease (aneurysm or acute dissection).⁷⁾ However, the cause of injury has not been clearly established. Our protocol for decreasing the risk of paraplegia after OSG now includes the following: identifying the artery of Adamkiewicz with use of multidetector row helical CT or magnetic resonance angiography before surgery; providing antegrade perfusion to the three cerebral artery branches; avoiding placement of the stent-graft distal to the artery of Adamkiewicz or Th8; careful deployment of the stent-graft delivery system with the aim of mitigating the risk of aortic wall injury and thromboembolism; and establishing blood return to the lower part of the body from the femoral artery by occluding the stent-graft with a balloon catheter during lower-body circulatory arrest. No cases of spinal cord ischemia have occurred in our patients since we began to use the blood-return procedure, but additional research on the pathogenesis of paraplegia after OSG is required.

Karck and Kamiya,¹²⁾ in a review of use of the frozen elephant trunk technique in 215 patients in eight series, reported a perioperative mortality rate of 7% (15 patients, including patients with acute aortic dissection). The overall perioperative mortality rate in our series was similar (6%), although the rate among our 44 patients who had elective surgery was lower (4%). So far, the long-term results of OSG in the treatment of TAA have not been studied extensively, but our rates of survival and freedom from an aortic event were comparable to those in other series (8-year survival, 53.7%–65.5%; 8-year aortic-event-free survival, 72.8%; and 8-year freedom from endoleak, 91.1%).^6,7)

Early and long-term results of OSG in patients treated for aortic dissection have been described in some recent reports.^11,13,14) According to Uchida, et al.,¹¹⁾ the incidence of patent false lumen on the proximal descending aorta after hybrid OSG is less than 7% and the distal reoperation rate is under 6% when endovascular reintervention is performed. Our series had six cases of type A and six of type B aortic dissection. In all 12 of these patients, entry closure and shrinkage of the false lumen were achieved. Thus, it is possible that OSG may be found to be a useful approach for treating aortic dissection.

The efficacy of OSG compared with other methods for treating extended TAA remains unknown. Clarification of the role of OSG in this application will require studies with larger numbers of patients and substantially longer follow-up times.¹⁵⁾ In our series, however, OSG using an MK device was effective initially and provided durable results. The flexibility and small diameter of the stent-graft allowed easy placement. The stent support over the full length of the graft may have prevented the device from bending and twisting on insertion and migrating in the long term. For the advantages, this technique could be an attractive treatment option for extended thoracic aortic aneurysms.

Acknowledgements

We thank Renée J. Robillard, MA, ELS, for editorial assistance and WL Gore & Associates for providing funding for this assistance.

Disclosure Statement

There is no conflict of interest.