HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Enio Buffolo Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Alves, C. M. R. Articles by Buffolo, E. Search for Related Content PubMed PubMed Citation Articles by Alves, C. M. R. Articles by Buffolo, E. Related Collections Great vessels

Ann Thorac Surg 2002;73:1143-1148
© 2002 The Society of Thoracic Surgeons

---

Original article: cardiovascular

Endovascular treatment of thoracic disease: patient selection and a proposal of a risk score

^a Federal University of Sao Paulo, Paulista School of Medicine, São Paulo, SP, Brazil

Accepted for publication December 18, 2001.

* Address reprint requests to Dr Alves, R. Simão Álvares, 527, ap. 63, Pinheiros, SP, São Paulo, Brazil CEP-05417-030
e-mail: cmralves{at}uol.com.br

	Abstract

Top Abstract Introduction Patients and methods Results Comment Acknowledgments References

 
Background. Although selection criteria and subgroup analysis are still in the early developmental stages, endovascular treatment of aortic disease has become an alternative to surgery for many patients.

Methods. From November 1996 to November 1999, 49 patients were treated with a self-expandable endoprosthesis at our institution. Most patients had acute aortic dissections. Thirteen of these patients did not follow the anatomic selection protocol. We retrospectively analyzed these patients to compare our numerical risk score (which includes clinical and anatomic criteria) between groups with or without success and between groups that followed the anatomic protocol (P) or did not follow the anatomic protocol (E [exception]).

Results. Success rates were similar in groups P and E, although mortality rates were higher in group E. Patients from group E had longer procedures and required multiple stents more frequently. The proposed risk score was able to differentiate between groups with or without success, as well as between groups P and E.

Conclusions. In order to reduce mortality and morbidity rates, careful selection criteria must be followed when treating patients endovascularly. Although it is time- consuming, using objective criteria can help select patients for endovascular treatment. We propose that patients with a risk score higher than 11 should only undergo percutaneous treatment when they have an unacceptably high surgical risk, and even so only after a detailed discussion of the risks.

	Introduction

Top Abstract Introduction Patients and methods Results Comment Acknowledgments References

 
Endovascular treatment of aneurysms and aortic dissections has been performed at various institutions, using different types of prostheses [1–4]. Even though many services related to these procedures are customized, the clinical and anatomic criteria for selecting patients have been monotonously similar. These criteria are defined in a clear way from the characteristics of each type of prosthesis (for example, the largest possible diameter) and, somewhat arbitrarily, from the subjective judgment of the physician performing the procedure as he or she takes into account the ability of the endoprosthesis to withstand tortuosities, to "land" in necks, and to adequately seal the entrance to the aneurysm. There seems to be the need to define objective criteria to assess the risk in undergoing this form of treatment, which despite being promising is still very new.

A numerical scoring system applied to candidates to minimally invasive treatment of aortic disease might help physicians decide whether or not to perform the procedure. As will be described in this article, a scoring system was developed by reviewing the outcomes of the initial cases in our experiment. A comparison of results between groups that either did (group P) or did not (group E) follow the anatomic protocol was also made.

	Patients and methods

Top Abstract Introduction Patients and methods Results Comment Acknowledgments References

 
Selection of patients
From November 1996 to November 1999, 49 endoprostheses were implanted through minimally invasive techniques in patients with descending thoracic aorta diseases of various etiologies (Table 1). The study protocol was approved by the institutional review board of the Sao Paulo Medical School and all patients signed a form giving their informed consent to undergo the procedure.

The clinical and anatomic indication criteria were as follows: (1) Presence of an acute or chronic type B aortic dissection (or its variants, intramural hematoma and penetrating ulcer) complicated by untreatable pain, rupture, or expansion plus a distance of at least 1 cm between the initial dissection and the left subclavian artery (LSA); (2) presence of a true aneurysm of the thoracic aorta with a diameter of at least 5.5 cm or a complicated (contained rupture, persistent pain, ischemia) aneurysm plus a distance of at least 1 cm between the initial dilation and the LSA; (3) presence of proximal and distal necks with diameters of up to 30 mm for the anchoring of the prosthesis; and (4) peripheral vascular access allowing the introduction of a stent-graft.

After the initial phase of the experiment, a group of 13 patients (26.5% of the population) was allowed percutaneous treatment despite not satisfying anatomic selection criteria (due to having a neck with a diameter greater than 30 mm, proximity to the subclavian artery, or a very tortuous aorta). In these cases, allowing percutaneous treatment was justified by an unacceptably high risk associated with surgery. These exception cases will be analyzed separately.

The self-expanding endoprosthesis used in all the cases is composed of a stainless steel cylindrical framework, made in Z segments, covered with polyester, with variable diameter and length (Braile Biomedics of Sao Jose do Rio Preto, Sao Paulo, Brazil; Fig 1). Such structure is compressed in a releasing catheter of polytetrafluoroethylene varying the final caliber of the device from 20F to 24F. The selected prosthesis had a diameter 10% to 20% larger than the diameter of the proximal neck.

View larger version (100K): [in this window] [in a new window]   Fig 1. Self-expanding endoprosthesis used in all cases (Braile Biomedics of Sao Jose do Rio Preto, Sao Paulo, Brazil).

Hypotension (mean arterial pressure of 50 to 60 mm Hg) with sodium nitroprusside was achieved before unloading the stent and immediately stopped after it.

Technical success was considered to have been achieved when, in the absence of death or surgical reversion, there was an immediate and significant fall in blood flow in the aneurysm or false lumen angiographically observed at first and confirmed in the 48-hour tomography scan. In the exception cases, technical success was considered to have been achieved when the intended treatment (such as covering thoracic tears) was successful, even if there was only a partial correction of the defect. Surgical conversion was advised in cases of serious vascular complications or in the persistence of significant flow in the aneurysm or false lumen.

Risk score definition
The score was developed after more than 3 years of experience in endovascular treatment of aortic diseases and it could not be done only on a statistical basis for each characteristic because of the small number of patients. Instead a combination of statistics and our own experience in patient selection was used. The weight for each possibility inside a criterion, although arbitrary, seemed appropriate. Then the score was retrospectively applied to the group here presented.

Seven clinical and angiographic characteristics were considered. For each characteristic a series of possibilities is listed, each with a different point value. In the case that more than one possibility describes the patient, the option with the highest numerical value is chosen. The sum of the points for all seven characteristics (anywhere from 0 to 23 points) defined a risk score whose mean was then compared in successful versus unsuccessful groups and also in groups that did versus those that did not follow the anatomic indication protocol. One patient was excluded from the score analysis because we could not measure all factors properly.

The seven characteristics and their respective scoring are as follows:

1. Criteria for complying with the anatomic protocol: fits all the criteria, 0 points (ie, proximal neck < 30 mm, single opening, not overly tortuous, perfect identification of the lesion); neck between 30 and 34 mm, 3 points; multiple openings to be covered, 3 points; overly tortuous access to the lesion, 3 points; inability of angiography to identify the lesion, 5 points; more than one criterion for anatomic exclusion, 5 points.
   
2. Etiology of the lesion: dissection variants, 0 points; type B dissection, 1 point; other etiologies, 2 points.
   
3. Treating the lesion requires passing part of the endoprosthesis through the aortic arch: no, 0 points; yes, 3 points.
   
4. Distance between the branching of the left subclavian artery and the beginning of the lesion: > 2 cm, 0 points; 0.5 cm and 2 cm, 2 points; < 0.5 cm, 4 points.
   
5. Angle between the arch and the descending thoracic aorta less than 90°: no, 0 points; yes, 2 points.
   
6. Presumed need for more than one stent to cover the lesion: no, 0 points; yes, 2 points.
   
7. Clinical variables: no sign of instability, 0 points; ruptured aneurysm, contained, 2 points; acute renal insufficiency secondary to the dissection, 2 points; recovery from cardiac and respiratory arrest or hypotension less than 90 mm Hg, 4 points; serious abdominal or lower limb ischemia, 5 points.
   

Statistical analysis
The qualitative variables were represented in absolute frequency (n) and relative frequency (%) and the quantitative variables were represented by their mean and standard deviation. We tested for correlation between qualitative variables with the ² test or with the Fisher test when the basic requirements for the ² test were not met.

When comparing two groups’ quantitative variables, we used Student’s t test for independent samples. When the variable was not normally distributed (especially when there was considerable variability within small samples), we compared the groups using the Mann-Whitney U test. In order to determine what scores indicate high or low risk for the procedure, we calculated the sensitivity and specificity for each score value found in the sample as a cutoff. We also plotted a receiver operating characteristic (ROC) curve to analyze the relationship between sensitivity and specificity. One chose a significance level of 0.05 ( = 5%). Descriptive levels (p) less than 0.05 were considered significant.

	Results

Top Abstract Introduction Patients and methods Results Comment Acknowledgments References

 
Forty-nine procedures were carried out in 49 patients with descending thoracic aorta disease. The mean age of the patients was 58.1 ± 13.5 years (range 18 to 80). The patients were mostly male (73.5%).

The procedure was considered a technical success in 43 patients (87.7%). Despite excellent technical results, 2 patients died in the hospital from unrelated causes: 1 from multiple system failure after cardiac and respiratory arrest and 1 from pulmonary embolism. Table 2 shows success rates based on the etiology of the lesion. The overall success rate was 83.7% (41 out of 49 patients).

When separately analyzing the patients presenting lesions requiring advancing part of the endoprosthesis through the aortic arch, we observed that there was success in 30 of 36 patients (83.3%) compared with 100% technical success rate (p = 0.175) for the 13 patients who did not require arch manipulation. In 35 of these 36 patients, the distance between the branching of the left subclavian artery and the opening of the dissection or the beginning of the lesion to be covered by the stent could be estimated by angiography. In 14 patients this distance was as much as 2 cm and in 21 patients it was more than 2 cm, with the technical success rates being 57.14% and 100%, respectively (p = 0.002). The endoprosthesis was released over the left subclavian artery in 4 patients, causing an acute or chronic obstruction of the artery without clinical consequences.

The mean duration of the procedure was 101 ± 58.8 minutes and the hospital stay was 7.7 ± 6 days.

Comparing results for protocol (P) group and protocol exception (E) group
As already mentioned some patients (n = 13) were selected for percutaneous treatment even though they did not completely satisfy the indication protocol. The groups were similar in regard to mean patient age and etiology of the lesions although there were predominantly atherosclerotic aneurysms in group E (30% versus 14%, p = 0.22).

We observed technical success rates of 88.9% for the 36 patients of group P and of 84.6% for the 13 patients of group E (p = 0.6495, odds ratio [OR] = 1.455, 95% confidence interval [CI]: 0.2 to 9).

The need for more than one stent was greater in group E (69.2% versus 22.2%, p = 0.002), as was the duration of the procedure (153.7 ± 70 versus 81.2 ± 40 minutes, p = 0.001) in this group.

Complications during hospital stay
There were 4 deaths during the hospital stay (8.16%), 2 from group P and 2 from group E (5.55% versus 15.4%, p = 0.284; OR = 3.09; 95% CI: 0.3 to 24.6). In the first 2 cases, death resulted from a secondary complication (multiple system failure after prolonged cardiorespiratory arrest that occurred before the procedure and pulmonary embolism on the 13th day after implanting the stent) despite the procedure itself being technically successful and tomography showing successful correction of the aneurysm. In the other 2 deaths, both from the group E, the procedure itself clearly played a role. In 1 case, after covering a large ruptured aneurysm with 3 stents in what was a very difficult and time-consuming procedure with extensive catheter and wire manipulation within the aortic arch, cardiac tamponade developed in the patient, who died from a dissection of the proximal aorta. In the other case, where angiography did not adequately show the opening of the dissection and its relation to the branching of the left subclavian artery, there was not any reduction of flow in the false lumen after releasing the endoprosthesis. After the procedure, there was persistent and severe ischemia to the lower limbs, which required urgent surgery. Death occurred 2 days after the procedure.

Vascular complications occurred in 4 patients: 1 brachial artery thrombosis, 1 femoral artery endarterectomy, 1 iatrogenic dissection of the femoral artery by the guidewire, and 1 deep venous thrombosis in the femoral vein.

There were 4 cases in which it was necessary to remove the prosthesis. In 1 case, releasing the endoprosthesis in a sharply angled aortic arch caused a partial and intermittent block to blood flow into the descending aorta (Fig 2). Three other patients needed to be sent to surgery because of persistent leakage (1 of these, who had lower limb ischemia, was already described).

View larger version (60K): [in this window] [in a new window]   Fig 2. The angle between arch and descending aorta may be an important factor for success. Mechanism of stent compression when an acute angle does not allow good apposition over the arch: (A) diastolic appearance; (B) systolic dislodgment of proximal end.

Despite the endoprostheses covering extensive parts of the descending thoracic aorta or being positioned in the region corresponding to vertebras T9 to T12, neither paraplegia nor paraparesis was observed in any of the patients.

Even though releasing multiple stents (n = 17) required more time-consuming procedures, it did not cause a significant increase in number of deaths or vascular complications (Table 3).

View larger version (11K): [in this window] [in a new window]   Fig 3. Receiver operating characteristic curve for score values as predictors of unsuccessful procedure, according to variables specified in the text.

	Comment

Top Abstract Introduction Patients and methods Results Comment Acknowledgments References

Retrospectively analyzing our group of patients with their diverse etiologies and anatomies it was possible to identify a series of characteristics that seem to indicate the likelihood of having a technically successful procedure. With this information it was possible to devise a scoring system. In this study, in which the first cases still used a prototype, we achieved an excellent success rate, similar to those reported in other studies [2, 3, 5, 6].

Even though unifactorial analysis shows that the only significant difference between the successful and unsuccessful groups was in the distance between the beginning of the lesion and the branching point of the left subclavian artery, the groups nevertheless had clearly distinguishable overall risk scores. The same was true concerning the protocol and protocol exception groups.

It is easy to explain why the other items were chosen to be included in the risk score. To begin with, both the need for arch manipulation and a sharp angle between the aortic arch and the descending aorta always represented an additional risk due to the possibility for dissection, stroke [5], or even the inability of the endoprosthesis delivery system to follow the course of the arch, allowing the stent to reach its target location. A sharp angle between aortic arch and descending aorta would in addition impair the accommodation of the stent in the arch, thus favoring leakage (which is without a doubt the method’s Achilles’ heel). Conversely, we observed a success rate of 100% in procedures carried out to treat thoracic artery disease where there was no need for arch manipulation, further corroborating this idea.

In all the studies reviewed it is recommended that the diameter of the prosthesis be about 20% larger than that of the target neck, allowing the stent to attach itself to a normal region of the aorta. In our study, even though the top limit for the diameter of candidate necks was determined based on the largest available prosthesis diameter, it is our conviction that necks with diameters larger than 32 mm might represent areas already damaged by the disease, facilitating the development of leakage in the long run.

Factors that are specific to the type of endoprosthesis being used (eg, ability to overcome tortuosities, maximum stent diameter, and caliber of the releasing device) may cause anatomic selection criteria to become more important or less important, changing their relative weights. Technologic advances in endoprosthesis development should have an even greater impact and may actually modify the criteria altogether.

The use of multiple stents causes, in addition to the obvious economic implications, the procedures to take more time, require a greater volume of contrast, more materials, and a potential to inflict more trauma to the femoral artery. In the study by Dake and associates [5], multiple stents were used in 33% of the patients (about the same as in our study), but this frequency may be as high as 70% [3]. The sequential placement of endoprostheses elevates both the immediate- and long-term risk because of the potential for leakage in the junctions between segments (a well-known problem in the endovascular treatment of abdominal aneurysms) [7]. However, the treatment of lengthy aortic segments in the region between T9 and T12 has been encouraging because the development of paraplegia seems to be less frequent than historically reported in surgical studies. At Stanford University, where studies with the largest populations to date have been conducted, paraplegia occurred in 3% of the group [8].

Treating patients who did not fit the anatomic protocol was in those patients’ best interest because their surgical risk was unacceptably high. Presenting the results from this group is valuable for two different but equally important reasons. First, in spite of the small number of patients in this study, we were able to observe a small difference in deaths related to the procedure (0% in group P and 15.4% in group E, p = 0.06, OR = 15.8, 95% CI: 0.7 to 355), which is very worrying and needs further analysis. However, success rates were similar between the two groups, allowing the method to be considered an option for these very sick patients who have a poor prognosis if left untreated. It is important to note that special techniques have also been used to lessen the risks for these complex cases such as carotid-subclavian bypass surgery before the procedure [8].

If one uses the risk score presented here (which we admit is still rather complex and needs prospective validation) to select patients using a score of 11 as the cutoff, one singles out a population with a high chance of success (92.5%) that is at low risk for complications. In patients who have a high risk score but also a very high surgical risk, the physician should discuss matters with the patient before considering the percutaneous procedure, as there was an important reduction in success rate for patients with scores greater than 11 points (62.5%, p = 0.0497; OR = 7.4; 95% CI: 1.1 to 47.2).

The inclusion of clinical criteria in the score is a consequence of our conviction that, in aortic disease, the clinical condition of the patient affects the outcome and when there are complications (especially in acute dissections), excluding the lesion from circulation is not always enough to achieve short-term recovery. For this reason, recommending the procedure, if compared with the high morbidity and mortality of surgery, should be in our opinion done sooner rather than later to also avoid the development of important anatomic distortion.

	Acknowledgments

Top Abstract Introduction Patients and methods Results Comment Acknowledgments References

 
Dr Alves received financial support from Conselho Nacional de Desenvolvimento Científico e Tecnologia (CNPq), Ministério da Ciência e Tecnologia.

	References

Top Abstract Introduction Patients and methods Results Comment Acknowledgments References

 

1. Dake M.D., Miller D.G., Semba C.P., Mitchell R.S., Walker P.J., Liddell R.P. Transluminal placement of endovascular stent-grafts for the treatment of descending thoracic aortic aneurysms. N Engl J Med 1994;331:1729-1734.[Medline]
2. Nienaber C.A., Fattori R., Lund G., et al. Nonsurgical reconstruction of thoracic aortic dissection by stent-graft placement. N Engl J Med 1999;340:1539-1545.[Medline]
3. Grabenwöger M., Hutschala D., Ehrlich M.P., et al. Thoracic aortic aneurysms treatment with endovascular self-expandable stents grafts. Ann Thorac Surg 2000;69:441-445.[Abstract/Free Full Text]
4. Fonseca J.H.P., Buffolo E., Carvalho A.C., et al. Utilização de endo-prótese auto expansível (stent) introduzida através da artéria femoral para tratamento de dissecção da aorta descendente. Arq Bras Cardiol 1998;70:389-392.[Medline]
5. Dake M.D., Miller D.G., Mitchell R.S., Semba C.P., Moore K.A., Sakai T. The "first generation" of endovascular stent-grafts for patients with aneurysms of the descending thoracic aorta. J Thorac Cardiovasc Surg 1998;116:689-704.[Abstract/Free Full Text]
6. Dake M.D., Kato N., Mitchell R.S., et al. Endovascular stent-graft placement for the treatment of acute aortic dissection. N Engl J Med 1999;340:1546-1552.[Medline]
7. Schlensak C., Blum U., Munzr T., Spillner G., Beyersdorf F. Three-year follow-up after endoluminal treatment for abdominal aortic aneurysm. Thorac Cardiovasc Surg 1998;45(suppl 1):177.
8. Fann J., Miller C. Endovascular treatment of descending thoracic aortic aneurysms and dissections. Surg Clin North Am 1999;79:551-574.[Medline]

This article has been cited by other articles:

				C. M. R. Alves, J. H. P. da Fonseca, J. A. M. de Souza, H. C. Kim, G. Esher, and E. Buffolo Endovascular Treatment of Type B Aortic Dissection: The Challenge of Late Success Ann. Thorac. Surg., May 1, 2009; 87(5): 1360 - 1365. [Abstract] [Full Text] [PDF]

				F. J. Criado, O. R. Abul-Khoudoud, G. S. Domer, C. McKendrick, M. Zuzga, N. S. Clark, K. Monaghan, and M. F. Barnatan Endovascular Repair of the Thoracic Aorta: Lessons Learned Ann. Thorac. Surg., September 1, 2005; 80(3): 857 - 863. [Abstract] [Full Text] [PDF]

				V. Rampoldi, S. Trimarchi, P. Righini, V. Tolva, and L. Inglese Open aortic surgical repair for left hemi-arch stent-graft failure Ann. Thorac. Surg., September 1, 2004; 78(3): 1075 - 1078. [Abstract] [Full Text] [PDF]

				G. Iannelli, F. Piscione, L. Di Tommaso, M. Monaco, M. Chiariello, and N. Spampinato Thoracic aortic emergencies: impact of endovascular surgery Ann. Thorac. Surg., February 1, 2004; 77(2): 591 - 596. [Abstract] [Full Text] [PDF]

				M. Hata, M. Shiono, T. Inoue, A. Sezai, T. Niino, N. Negishi, and Y. Sezai Optimal treatment of type B acute aortic dissection: long-term medical follow-up results Ann. Thorac. Surg., June 1, 2003; 75(6): 1781 - 1784. [Abstract] [Full Text] [PDF]

				E. Buffolo, J. H. P. da Fonseca, J. A. M. de Souza, and C. M. R. Alves Revolutionary treatment of aneurysms and dissections of descending aorta: the endovascular approach Ann. Thorac. Surg., November 1, 2002; 74(5): S1815 - S1817. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Enio Buffolo Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Alves, C. M. R. Articles by Buffolo, E. Search for Related Content PubMed PubMed Citation Articles by Alves, C. M. R. Articles by Buffolo, E. Related Collections Great vessels