|
|
||||||||
J Thorac Cardiovasc Surg 2006;132:1030-1036
© 2006 The American Association for Thoracic Surgery
Surgery for Acquired Cardiovascular Disease |
a Department of Thoracic and Cardiovascular Surgery, Rangueil University Hospital, Toulouse, France
b Department of Radiology, Rangueil University Hospital, Toulouse, France
e Department of Cardiology, Rangueil University Hospital, Toulouse, France
c Department of Epidemiology and Biostatistics, Toulouse University Hospital, Toulouse, France
d Department of Cardiology, Montauban General Hospital, Montauban, France.
Received for publication March 4, 2006; revisions received June 24, 2006; accepted for publication July 7, 2006. * Address for reprints: Hervé Rousseau, MD, Radiology Department, Rangueil University Hospital, TSA 50032, 1 Avenue Jean Poulhès, 31059 Toulouse, Cedex 9, France. (Email: rousseau.h{at}chu-toulouse.fr).
| Abstract |
|---|
|
|
|---|
METHODS: Between 1996 and 2005, 45 patients (mean age, 68 ± 11 years) with aneurysms of the descending thoracic aorta underwent endovascular repair. Aortic dissections, penetrating ulcers, and traumatisms were excluded. The mean follow-up was 24.7 ± 21.6 months (maximum, 6.7 years).
RESULTS: No patients died, and no conversion to surgical intervention was required during the procedures. Three (6.7%) patients died during the first month, and 6 (14.7%) died later on. The main complications were strokes (13.3%), vascular access complications (8.9%), aortic complications (6.6%), paraplegia (4.4%), and sudden deaths (4.4%). Nineteen (42%) primary endoleaks were encountered: 3 required reinterventions, 15 spontaneously thrombosed, and 1 patient died. Except for 2 sudden unexplained deaths, no aortic complications were observed after 1 month. Actuarial survival estimates at 1, 3, and 5 years were 87.6% ± 5.3%, 76.9% ± 7.4%, and 70.6% ± 9.2%, respectively. Actuarial freedom from death related to the treated aortic disease was 94.3% ± 4.0%, 94.3% ± 4.0%, and 86.4% ± 8.4% at 1, 3, and 5 years, respectively. Aspirin status of greater than 3 (P = .005), high aortic diameter (P = .007), and long covered lengths (P = .02) were determinant for mortality. Actuarial freedom from complication was 62.6% ± 7.7%, 58.9% ± 8.1%, and 58.9% ± 8.1% at 1, 3, and 5 years, respectively. The location of the aneurysm (P = .05) and a high aortic diameter (P = .04) were both determinants for endoleaks.
CONCLUSIONS: Stent grafting of atherosclerotic aneurysm of the thoracic descending aorta is safe and effective. Further studies are mandatory to determine the most relevant indications and the long-term efficacy of such treatment.
| Introduction |
|---|
|
|
|---|
The first stent-grafting procedure in the human thoracic aorta was allotted to Volodos and colleagues in 1991.2
It quickly offered the promise of a new and less-invasive method of achieving aneurysm exclusion.3
A restricted number of groups demonstrated technical feasibility with perioperative morbidity and mortality rates lower than those encountered after conventional surgical treatment.4-12
Even so, the endovascular repair of aneurysms of the descending thoracic aorta is less widespread than procedures carried out on the infrarenal aorta. The long-term results are thus not well known.
The objective of this study is to report the midterm results of endovascular repair of aneurysms of the descending thoracic aorta with the exclusive use of second-generation stent grafts.
| Patients and Methods |
|---|
|
|
|---|
|
|
Follow-up
Follow up consisted of standard chest roentgenography, transesophageal echocardiography, and injected spiral scanner or magnetic resonance before discharge and on the 3rd, 6th, 12th, and 18th months. Standard chest roentgenography and injected spiral scanner or magnetic resonance were performed every subsequent year. One patient was lost from follow-up (2.2%). Mean follow-up was 24.7 ± 21.6 months (range, 6 months-6.7 years).
Definitions
The definitions used in this study are described in the literature.13
Primary technical success was defined as a complete exclusion of the aneurysm without an additional procedure and without perioperative death or complication. Secondary technical success was defined as a complete exclusion of the aneurysm after an additional endovascular procedure or after spontaneous thrombosis of primary endoleak within the first month. Endoleak was defined as the angiographic, tomodensitometric, or echocardiographic description of blood flow apart from the stent graft. Endoleaks were then classified according to the literature.14,15
Any perioperative deaths, any primary or secondary endoleaks, any failures of the endovascular device, any reinterventions, any deaths related to an aortic rupture, or any sudden unexplained late deaths were considered as failures of the treatment.
Statistical Analysis
All the statistical studies were carried out with the STATA program (Stata Corp, College Station, Tex). The continuous variables were expressed as means ± standard deviations.
The evaluation of actuarial survivals was carried out according to the Kaplan-Meier nonparametric method. The actuarial rates of survival were reported as 1 ± standard deviation.
The multivariate study was performed by using the log-rank test, and the factors analyzed as potential predictors of adverse events were the age and sex of the patient, aspirin status, comorbidities (hypertension, diabetes mellitus, previous cerebrovascular accident, coronary disease, renal failure, and chronic pulmonary obstructive disease), emergency operation, location of the aneurysm, locations of the proximal and distal landing zones and their diameters, maximal diameter of the aneurysm, number of stent grafts used, covered length, and coverage of the T9-L1 intercostal arteries.
| Results |
|---|
|
|
|---|
Early Outcome
Mortality
Three patients died during the 30 days after the procedure. The causes of the deaths were 2 pneumoniae (days 6 and 28) and 1 aortoesophageal fistula, which was a consequence of the aneurysm before stent grafting (day 30). The perioperative death rate was 6.7%. The death rate related to the treated aortic disease was 2.2%. The death rate at 30 days of the patients treated in emergency situations was 12.5%. It was 5.4% for the patients electively treated (P > .05).
Morbidity
In the first 30 days after the intervention, 20 complications were encountered in 15 patients. Five patients presented with vascular complications: 2 with ischemia of the left upper limb, 2 with pseudoaneurysms of the left brachial artery, and 1 with rupture of the iliac artery. Nine patients presented with neurological complications. Cerebrovascular accidents were encountered in 6 patients, and this was the main complication (13.3%). Of the 6 patients with their left subclavian artery covered, one patient presented with an ischemic left frontoparietal cerebrovascular accident. In all the other cases, the encephalic injuries were diffusely positioned intracerebrally and were also consistent with the manipulation of intravascular devices in an atherosclerotic aortic arch. Two (4.4%) patients presented with paraplegia, and one of them had delayed paraplegia caused by a rupture of an external iliac artery with collapse. One patient presented with a transient left recurrent laryngeal nerve paralysis, which was identified during postoperative fiberoptic laryngeal examination. Two patients presented with severe respiratory complications and died from pneumonia.
Endoleaks, aortic complications, and reinterventions
Nineteen (42.2%) primary endoleaks were encountered: thirteen (28.8%) type I endoleaks, 1 (2.2%) type II endoleak, and 5 (11.1%) type III endoleaks. The primary technical success rate was 58.8%. Two secondary endoleaks (1 type I and 1 type III) were observed, both within the first 30 days (4.4%).
Aortic complications occurred in 4 patients. These complications were always associated with primary or secondary endoleaks. One contained rupture was associated with a secondary type III endoleak (day 23; Figure 1, A and B). Three fistulae were associated with type I endoleak. One patient died on postoperative day 30 with a de novo aortobronchial fistula (primary type I endoleak). One aortoesophageal fistula (primary type I endoleak) and 1 aortobronchial fistula (secondary type I endoleak), which were both consequences of the aneurysm before stent grafting, recurred on postoperative days 9 and 31, respectively, and required reintervention.
|
Shrinkage of the aneurysmal sac
During the first month, the aneurysmal diameter increased in 3 cases. In 2 cases the increase in diameter was transitory (5 and 10 mm) without detectable endoleak, and in 1 case an increase in diameter of 30 mm was observed. This latter case was associated with a primary type I endoleak in a patient with 2 major postoperative complications (cerebrovascular accident and renal insufficiency), thus preventing the performance of reintervention. The patient died on postoperative day 30, with no reintervention having been performed. In all the other cases, the aneurysms shrank. The mean shrinkage of the aneurysmal sac was 12.8 ± 10.3 mm at the end of the first month.
Midterm Outcome
No device failure, no migration, no kink or twist, and no secondary endoleak were observed beyond the 30th day after the procedure. No aortic reintervention was required after the second month. Six patients died beyond the 30th day after the procedure. The causes of the deaths were renal insufficiency 62 days after the procedure, digestive hemorrhage in a cirrhotic patient 1.1 years after the procedure, myocardial infarction 1.8 years after the procedure, and cerebrovascular accident 3.1 years after the procedure. Two patients died suddenly 1 and 1.6 years after the procedure without any obvious cause having been highlighted and without any autopsy having been performed.
Survival Analysis
Actuarial survival estimates 1, 3, and 5 years after the procedure were 87.6% ± 5.3%, 76.9% ± 7.4%, and 70.6% ± 9.2%, respectively (Figure 2, A). More specifically, actuarial freedom from death related to the treated aortic disease was 94.3% ± 4.0%, 94.3% ± 4.0%, and 86.4% ± 8.4% at 1, 3, and 5 years, respectively (Figure 2, B). Actuarial freedom from complication was 62.6% ± 7.7%, 58.9% ± 8.1%, and 58.9% ± 8.1% at 1, 3, and 5 years, respectively (Figure 2, C).
|
Multivariate Analysis
We identified aspirin score equal to or greater than 3 (P = .005), high diameter of the aneurysms (P = .007), more than 1 stent graft deployed (P = .0001), and long covered lengths (P = .02) to portend a significantly higher likelihood of late death.
We identified the location of the aneurysm (P = .05) and a high diameter of the aneurysm (P = .04) to portend a significantly higher likelihood of secondary endoleak. We identified the location of the aneurysm (P = .01) and high diameter of the aneurysm (P = .04) to portend a significantly higher likelihood of treatment failure (P = .01).
| Discussion |
|---|
|
|
|---|
Although Dake and coworkers,20
using homemade devices, reported 58% femoral access and up to 30% aortic approach, we were able to introduce the stent graft through an infrainguinal iliofemoral approach in all the cases. Nevertheless, rupture of the external iliac artery was encountered in 2 patients. In one case this rupture was associated with secondary paraplegia because of a hemodynamic collapse, thus underlining that these vascular complications might have severe consequences.
Accurate placement of the stent graft in the arch is one of the most challenging technical feats, and perioperative stent graft migration is reported to occur in 2% to 20% of cases.21,22
We avoided this problem with lower blood pressure during deployment, and we did not convert any endovascular procedure to open surgical intervention, which is a major improvement compared with the 4% and 12% conversion rate, respectively, reported by Dake23
and Greenberg and associates.7
This can also probably be explained by the technical improvements of the devices used, allowing partial deployment and correct repositioning before complete dropping into place of the devices.
Early mortality is the most important criterion in appraising the results of thoracic aortic aneurysm treatment and in comparing endovascular repair with conventional open surgery. According to the literature, the 30-day mortality rates range between 0% and 15%.4,8,18,19,21,22,24-27
We report a 30-day mortality rate of 6.6%. Two patients died of severe pneumonopathy (keep in mind that chronic pulmonary disease was the main comorbidity), and 1 patient presented with an aortic-related 30-day death.
Paraplegia is one of the most feared complications after repair of thoracic aneurysms. Although the endovascular repair of thoracic aortic aneurysms shows a promising reduction in operative morbidity compared with open surgery, the risk of spinal cord ischemia still exists. According to the literature, paraplegia occurs in 0% to 6.5% of cases.4,10,18,21,25,27-29
Two (4.4%) of our patients presented with paraplegia, one immediately. The other patient presented with secondary paraplegia after delayed hemodynamic collapse. Despite appropriate medical treatment, including cerebrospinal fluid drainage, these patients did not recover from their deficit. Because of a limited number of events, we were unable to identify any risk factors for paraplegia.
As reported by other authors,4,10,18,21,25,27-29
the main postoperative complication of endovascular repair of aneurysms of the descending thoracic aorta was cerebrovascular accident (13.3%). Because the encephalic injuries were diffusely positioned, the occurrence of a cerebrovascular accident can easily be explained by the manipulation of wires and catheters in an atherosclerotic aortic arch. We try to minimize the risk of embolic stroke by limiting the manipulation within the aortic arch, particularly when severe atherosclerotic aortic arches are identified: we avoid handling the guides as much as possible at the level of the aortic valve and of the arch, a single guide wire is placed proximally to the aneurysm, we always maintain the rigid part of the guide wire upstream of the dropping device to decrease friction on the wall, and minimal progression of the undeployed stent graft is performed across the arch.
Primary endoleak rates are similar to those encountered in the infrarenal aorta (4.2% to 53%).23
Forty-two percent of our patients experienced primary endoleak, identified at the end of the procedure by means of angiography, transesophageal echocardiography, or both. We observed type I (28.8%) and type III (11.1%) endoleaks and 1 type II endoleak related to an intercostal artery. Contrary to Greenberg and associates,7
we do not conclude that the presence of a primary endoleak should always be treated aggressively because most of the primary type I endoleaks (7/13) spontaneously thrombosed during the first month. Of 13 primary type I endoleaks, 7 spontaneously healed during the first month, and 4 required reintervention with a good result. One patient with an aortoesophageal fistula died, and another with an increase of diameter died because of another complication preventing the performance of reintervention. In fact, numbers of type I endoleaks are minor (mostly when identified on transesophageal echocardiography) and spontaneously heal, provided that the choice of the diameter of the stent graft and its positioning are optimal. In these cases we are content to model the prosthesis with a balloon to achieve optimal proximal seal. In case of an evident error in the diameter or in the positioning of the device, it is necessary to consider the use of a proximal long extension with a higher diameter (and not a cuff, which would not present a sufficient overlapping). If the endoleak persists, even remotely, it would be necessary to perform reintervention. Two patients experienced secondary endoleaks, 1 type I and 1 type III, during the first postoperative 30 days. Concerning the secondary type I endoleak, the proximal neck was probably too short, whereas no technical failure was identified in the secondary type III endoleak. Both of the secondary endoleaks were associated with aortic complications. No secondary endoleak was observed beyond postoperative day 30.
Despite a longer follow-up time than most of the references in the literature, we did not observe any increased long-term complication rate. Six patients died, but only 2 deaths were potentially related to the treated aortic disease (2 unexplained sudden deaths).
| Limitations of Our Study and Current Prospects |
|---|
|
|
|---|
| Conclusions |
|---|
|
|
|---|
| Acknowledgments |
|---|
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Marui, T. Mochizuki, T. Koyama, and N. Mitsui Degree of fusiform dilatation of the proximal descending aorta in type B acute aortic dissection can predict late aortic events. J. Thorac. Cardiovasc. Surg., November 1, 2007; 134(5): 1163 - 1170. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. G. Bakaeen, J. S. Coselli, S. A. LeMaire, and J. Huh Continued Aortic Aneurysmal Expansion After Thoracic Endovascular Stent-Grafting Ann. Thorac. Surg., September 1, 2007; 84(3): 1007 - 1008. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |