This article was published in an Elsevier journal. The attached copyis furnished to the author for non-commercial research and

education use, including for instruction at the author’s institution,sharing with colleagues and providing to institution administration.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

European Journal of Radiology 64 (2007) 65–72

Endovascular stent-graft treatment of thoracic aortic syndromes:A 7-year experience

Giovanni Dialetto a, Alfonso Reginelli b, Marcella Cerrato b, Giovanni Rossi c,Franco Enrico Covino a, Sabrina Manduca a, Francesco Lassandro c,∗

a Department of Cardiothoracic and Respiratory Sciences, Second University of Naples, V. Monaldi Hospital, Napoli, Italyb Department of Radiology, Second University of Naples, Napoli, Italy

c Department of Radiology, Monaldi Hospital, Napoli, Italy

Received 15 June 2007; accepted 25 June 2007

Abstract

Thoracic aortic diseases (TAD) are relatively frequent conditions associated with high mortality. Recently, several reports have demonstratedthe safety and efficacy of endovascular stent-graft (EVG) placement for TAD as an alternative to open surgery.

We report our experience in management of thoracic aortic syndrome on 56 consecutive patients with TAD that underwent endovascularstent-graft repair.

MDCT angiography was used in all patients to provide preprocedure evaluation and measurements. In particular it is necessary to evaluate theproximal and distal landing zones of the stent-graft.

All EVGs in our series were placed successfully. Conversion to open surgery was never required. Six patients (10.7%) died early after thestent-graft deployment. During follow-up four more patients died. The endoleak rate was 16.7% (no. 10 pt). We did not observe any case ofparaplegia.

The present study shows the efficacy of EVG in the long-term follow-up, with an overall survival of 82.1%, which is comparable to that reportedin recent studies. In conclusion this technique is emerging as an alternative approach in the treatment of TAD because this approach offers a lessinvasive therapeutic option to standard surgical techniques, even in patients who have associated diseases that make them poor surgical candidates.© 2007 Elsevier Ireland Ltd. All rights reserved.

Keywords: Thorax; Aorta; Endovascular

1. Introduction

Thoracic aortic diseases (TAD) are relatively frequent con-ditions associated with high mortality. They include aneurysms(Fig. 1), dissection, rupture, traumatic pseudoaneurysms, intra-mural hematoma and penetrating atherosclerotic ulcers (Fig. 2)[1,2].

This complex group of disease shares relatively similar clin-ical presentations and imaging protocols [3,4]. Surgical graftreplacement has been until recently the only effective treatmentmethod in these conditions. Despite significant improvement inanaesthesiological and surgical techniques and in perioperativecare, transthoracic aortic surgery carries significant mortality

∗ Corresponding author at: via Stanzione 18, 80129 Napoli, Italy.E-mail address: f.lassandro@tiscali.it (F. Lassandro).

(5–15% in elective cases and up to 50% in emergency situa-tions) and risks of serious complications such as paraplegia,renal failure and need of prolonged ventilatory support [1,5].Recently, several reports have demonstrated the safety and effi-cacy of endovascular stent-graft (EVG) placement for TAD asan alternative to open surgery [6,7].

In clinical practice the imaging procedure most fre-quently used in the diagnostic assessment of these disease ismultidetector-row computed tomography (MDTC) [8]. MDCTallows an accurate preprocedure evaluation, essential to selectpatients for endovascular treatment, to choose the stent-graftdevices, and to plan the intervention (Fig. 1) [9]. In this paperwe present our 7-year clinical experience in management ofthoracic aortic syndrome.

We report the procedure details of endovascular treatmentsof TAD including preprocedure imaging, the endovascular pro-cedure and follow-up. We discuss and illustrate the imaging

0720-048X/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.ejrad.2007.06.019

Author's personal copy

66 G. Dialetto et al. / European Journal of Radiology 64 (2007) 65–72

Fig. 1. Degenerative aneurism with aortic rupture at the isthmus. CT scan: mediastinal hemorrhage with contrast media extravasation (a). Digital subtractionangiography (DSA): Contrast extravasation from the aortic rupture (b). An endovascular procedure was performed. EVG was positioned in the descending aorta withcomplete exclusion of the aneurysmal sac. At DSA the EVG was correctly positioned, contrast extravasation was no more apparent (c). On the postoperative CTscan EVG resulted correctly positioned in the descending aorta with complete exclusion of the aneurysm. In the mediastinum there was still blood collection, but nocontrast extravasation was detectable (d). The patient did not experience any problem on follow-up.

findings and techniques in aortic aneurysms, dissections, trau-matic pseudoaneurysms, intramural hematoma, and penetratingatherosclerotic ulcers.

2. Materials and methods

2.1. Patients population

From April 2000 to March 2007, 56 consecutive patients withTAD underwent endovascular stent-graft repair at “V. Monaldi”Hospital of Naples, Italy. They were seven women (age range,39–73 years; mean age, 67.4 years) and 49 men (age range,17–83 years; mean age, 64.3 years). Prior informed consent forthe procedure was obtained from every patient according to aprotocol approved by the institutional review board.

Cases included degenerative aneurysm in 23 patients (Fig. 1),traumatic pseudoaneurysm in eight and non-traumatic pseu-

doaneurysms in 2 patients, thoracic aortic dissection in 23,comprehensive of four intramural hematoma and one penetrat-ing ulcer (Fig. 2). Thirty-one patients (55%) were treated inemergency conditions.

Enrolment criteria: selection criteria for endovascular treat-ment are reported in Table 1.

2.2. Preprocedure imaging

MDCT angiography was used in all patients to providepreprocedure evaluation and measurements (Fig. 1). At our insti-tution, the standardized protocol consisted of an unenhanced CTstudy and a contrast-enhanced CT study from the thoracic inletto the femoral artery bifurcation.

We used a single slice CT (Toshiba Xpress/GX) in 14/60procedures, scan parameters: collimation 3 mm, reconstructionslice thickness 3 mm, pitch 1.5, 1 s rotation. After January

Author's personal copy

G. Dialetto et al. / European Journal of Radiology 64 (2007) 65–72 67

Fig. 2. A penetrating ulcer of the descending thoracic aorta observed at DSA (a) and successively at intracardiac echocardiography (arrow) (b). During the endovascularprocedure graft distal sealing was observed at intracardiac echocardiography (c) and the complete exclusion of the penetrating ulcer was confirmed at angiography(d). The follow-up was uneventful.

Table 1

Inclusion criteria1. Descending aorta aneurysm, type B aortic dissections, type B intramural hematoma, descending aorta penetrating ulcer, descending aorta traumatic or

non-traumatic pseudoaneurysm with indications for surgery.2. Lack of vascular anomalies that could interfere with stent-graft insertion and

deployment.3. Presence of a suitable landing zone proximally and distally to the lesion.4. Diameter of the aortic segment immediately proximal to aortic lesion ≤40 mm.5. Patients with overall stable hemodynamic parameters.

Patients excluded from the study:1. Patients with ongoing pregnancy or malignancy.2. Complex dissection and course of the false lumen at preoperative angiogram with impossibility to correctly stent the true lumen.3. Severe respiratory failure that might preclude a possible thoracotomy.4. History of myocardial infarction or cerebrovascular accident in the last 6 weeks.5. Abuse of drugs in the last 6 months.

Author's personal copy

68 G. Dialetto et al. / European Journal of Radiology 64 (2007) 65–72

2005 patients (16/60 procedures) were studied with a sixteen-slice MDCT (Philips Brilliance) with the following scanningparameters: collimation 16 mm × 1.5 mm, reconstruction slicethickness 2 mm, 0.5 s rotation, pitch 0.938 were used. Non-ioniccontrast media at a concentration of 370 mg/I was injected usinga power injector at a flow rate of 3–4 ml/s. Contrast media vol-ume was defined on the basis of patient’s weight (2 ml/kg of bodyweight) up to a maximum total volume of 150 ml. The acquisi-tion started when a peak of aortic enhancement measured witha specific program was reached (bolus tracking).

We did not repeat CT in patients having already performed aCT scan elsewhere (30/60 procedures), when all measurementswere possible and in particular when CT scans were performedwith helical technique and with at last a 3 mm collimation param-eter.

On unenhanced examination calcifications or fresh blood inaortic wall were searched on.

On contrast-enhanced CT were evaluated: (a) the location ofthe aortic lesion and its exact anatomy and morphology; (b) thepresence of mural thrombus or calcifications; (c) the exact intra-luminal measurements of the diameters of the aortic lesion andthe aortic segments proximal and distal to it; (d) the relation-ships of the aneurysm neck with adjacent aortic branch vessels,mainly the left subclavian artery (LSA) and celiac trunk; and (e)the anatomy of the abdominal aorta, the extension of the lesionand the size and tortuosity of femoral and iliac vessel [10]. Allthese findings were correlated to the other available radiolog-ical data and with medical follow-up. Particular attention wasdevoted to aortic wall status at the proximal and distal neck ofthe aneurysm to identify mural thrombus or calcifications thatcould interfere with sealing of the aneurysm.

In case of aortic dissection informations included identifica-tion of entry and re-entry sites, relationships among true andfalse lumen with visceral and femoral vessels, true and falselumen dimensions, and diameter of distal aortic arch, as well asdiameter and extension of the dissection to the femoral and iliacarteries.

In eight patients (14.3%) digital angiography was performedbefore the implantation of EVG to better study entry tear in caseof aortic dissection and the course of the true and false lumen.

In 25 patients (44,6%) trans-oesophageal echography (TEE)was performed before the implantation of EVG to get more infor-mation about the wall, the tears in case of dissection, and thethrombus.

3. Endovascular procedure

3.1. Stent-graft systems

The dimensions of the stent-graft used were determined onthe basis of contrast-enhanced CT and TEE images. Stent-grafts were selected according to the aortic diameter and lengthof the lesion [11]. For optimal fixation, all stent-grafts wereoversized in diameter compared with the diameter of the prox-imal and distal necks of the lesion, by 15–20% in cases ofatherosclerotic aneurysms and by 10–15% in all other cases[10]. We used two different first generation EVGs: 34 (44.1%)

TalentTM(Medtronic, World Medical Manufacturing Corp., Sun-rise, FL) and 14 (18.2%) Excluder® (Gore, Sunnyvale, CA).Furthermore, we used as a second generation EVG: in 22(28.6%) ValiantTM(Medtronic, World Medical ManufacturingCorp., Sunrise, FL) and 7 (9.1%). TAG (Gore, Sunnyvale, CA)for a total of 77 EVGs placed in 56 patients. Technical featuresand deployment of both systems have already been described indetail. EVG placement procedures were performed in the oper-ating room with patients receiving general anaesthesia. TEEwas performed in all patients but one, in which intracardiacechocardiography was performed. The common femoral arterywas surgically exposed and controlled proximally and distally.If the femoral artery appeared inadequate, the iliac artery wasused for vascular access.

An angiography catheter was inserted through the right and/orleft brachial artery, and intermittent contrast material injectionswere made before and during stent-graft deployment. When thelesion was too close to the left subclavian artery (LSA), anadditional catheter was advanced in the ascending aorta via apercutaneous left brachial approach and used as an adequatelandmark for the LSA origin for delivery.

A hydrophilic guidewire was placed in the aortic arch underfluoroscopic guidance via the elected vascular access. It wasthen exchanged for a stiff guidewire. 100 IU/kg of heparin wereadministered in each case. The delivery system was passed overthe stiff guidewire and positioned at the proximal end of theaortic abnormality. The exact placement site was selected onthe basis of CT, angiographic and TEE informations, aortic wallstatus and diameter at the neck sites were considered. Duringrelease of the device, the systolic arterial blood pressure waslowered to 75 mmHg. When needed, a balloon catheter wasinflated to achieve full expansion and to anchor the EVG to theaortic wall. In some cases additional devices were deployed toensure disease exclusion. Final angiography and TEE or intrac-ardiac echocardiography were performed to confirm properEVG placement and complete disease exclusion and to verify thepresence of correct perfusion through the graft without perigraftleakage (Figs. 1 and 2). After removal of the delivery system,the vascular access was closed by standard surgical closure tech-niques. No further anticoagulation was administered.

A further element to consider is a short proximal neck closeto the left subclavian artery. We favoured the exclusion of theartery itself in all those situations (five patients) in which a moreproximal treatment of the aortic lesion was feasible. In two casesthe left subclavian artery was prior grafted to the left carotid[12,13].

4. Follow-up

Our follow-up protocol consisted of CT performed beforedischarge. At 3, 6 and 12 months postoperatively, and annu-ally thereafter or whenever complications were suspected TEEand/or MDCT were performed.

The procedure was considered technically successful whenthe primary entry tear of the aortic dissection or the(pseudo)aneurysm was completely excluded from the circula-

Author's personal copy

G. Dialetto et al. / European Journal of Radiology 64 (2007) 65–72 69

tion and no perigraft leakage was detected on follow-up CT/TEE.Obliteration and thrombosis of the aneurysmal sac or pseudoa-neurysm or the false lumen were examined in patients withdegenerative aneurysm, traumatic aortic rupture or aortic dis-section.

Early mortality and morbidity included events occurringwithin 30 days after stent-graft deployment, either in hospitalor after discharge. Informations about patients were obtainedboth from retrospective chart review and by contacting patientsor their primary physicians.

The follow-up range was 0.5 to 64 months, its mean durationwas 24.9 ± 21.6 months and it was complete in 53 pts (94.6%).

According to accepted nomenclature [14,15], we classifiedendoleaks as follows: type 1 proximal (1a) or distal (1b), in casesof involvement of the proximal or the distal attachment site ofthe endovascular graft to the vascular wall; type 2, in cases ofthe presence of liquid blood originating from a tributary to theaortic lumen excluded by the graft with or without outflow to asecond tributary; type 3, in cases of an endoleak arising from adefect of the graft itself as a laceration, a hole, or a componentseparation of modular grafts (Fig. 3); and type 4, in cases of anendoleak caused by stent macroporosity.

Endoleak was also classified as primary (diagnosed within30 days of endovascular repair) or secondary (diagnosed morethan 30 days after intervention). All complications occur-ring while the patient remained in the hospital or later, wererecorded.

5. Results

In all cases, the EVG was placed successfully; neither pro-cedures aborted because of access difficulty, nor conversion tostandard open repair were necessary. Overall, the mean dura-tion of the procedure was 119 ± 75 min (range 45–240). Medianblood loss was 254 ml (range 50–1200 ml). The average totalamount of contrast was 150 ± 50 ml (range 50–340 ml). A totalof 77 EVGs were implanted through 60 procedures for a total of56 patients: in 45 procedures we used a single EVG, in 13 twoEVGs, and in 2 three EVGs. The LSA was over-stented in 5 pts.In patients examined until 2002 we performed a carotid to sub-clavian by-pass when the graft was to close to the left subclavianartery (2 pts). Since 2003 we used to perform a by-pass from thecarotid artery to the left subclavian artery only when a criticalcerebral circulation was detected on MRI study performed incase of short proximal neck (12 pts) and the procedure was notperformed in emergency.

In conclusion in two cases (3.6%) we performed a carotid-subclavian by-pass, axillary-bifemoral by-pass (n = 1) andsovrapubic femoro-femoral by-pass (n = 2).

5.1. Mortality

Six patients (10.7%) died early after the stent-graft deploy-ment. The causes of death were multi organ failure (politrauma)(n = 4), aortic rupture (n = 1), neurological sequelae (n = 1).

During follow-up four more patients died. The causes of deathwere prostate cancer (n = 1), sudden death for suspected aor-

tic rupture (n = 2), acute respiratory failure related to chronicobstructive lung disease (n = 1).

5.2. Neurological sequelae

In our series there was no paraparesis or paraplegia due toischemia of the spinal cord. One patient developed cerebrovas-cular events after the operation.

5.3. Endoleaks

Overall, the endoleak rate was 16.7% (no. 10 pt/60 proce-dures) comprehensive of two cases with flow in the false lumenfrom accessory tears.

Conversion to open surgery was never required.Early endoleaks were detected before discharge in 6/60 pro-

cedures (10%): five type 1a endoleak. In 1 patient with typeB aortic dissection and 1a endoleak we observed aneurysmaticdilatation of the descending aorta, but the pts refused a new treat-ment to cover the proximal tract stent and near aorta. In anotherpts with type B aortic dissection and type 1a endoleak a surgicalintimal fenestration in the abdominal aorta was performed todecompress the false lumen.

In one patient with type B aortic dissection there was a patentfalse lumen due to a tear located 2 cm downstream from the distalsealing site very near to the celiac trunk. This patient showed aprogressive dilatation in the follow-up.

Late endoleaks occurred in 4/60 procedures (6,7%): 1 type 1bendoleak in a patient with degenerative aneurysm that refusednew treatment and died 6 months after the first implantation,1 type 1a treated by additional EVG, and 1 type 3, treated byadditional EVG (Fig. 3). Finally in one patient with type B aor-tic dissection in the follow-up a new entry tear was detectedbefore the EVG, thus also this patient was treated by additionalEVG.

6. Discussion

The surgical repair of the descending aorta has the risks ofhigh morbidity and mortality [4,12,16] . The population affectedoften is elderly, and these patients frequently have associateddiseases that can determine a high surgical risk [17–19]. In thelargest surgical series, the surgical mortality ranged from 5 to20%, and postoperative renal and pulmonary insufficiency fre-quently resulted in a prolonged hospital stay [20]. Moreover,despite many attempts to prevent ischemic damage to the spinalcord, paraplegia continues to be a devastating complicationassociated with this type of surgery. The use of endovascu-lar stent-grafts in the treatment of diseases of the thoracicaorta represents a promising alternative to open surgical repair.Endovascular techniques were initially applied in the treatmentof abdominal aortic aneurysms (10), and this application stim-ulated investigation into the feasibility of endovascular repairof the thoracic aorta. Initial clinical reports and the results ofthe prospective trial of the Stanford group indicated lower mor-bidity and mortality of endovascular treatment in comparisonwith findings at conventional surgery [9,21]. Findings in our

Author's personal copy

70 G. Dialetto et al. / European Journal of Radiology 64 (2007) 65–72

Fig. 3. Rupture of a descending thoracic aortic aneurysm treated with EVG. At 9 months follow-up the EVG was correctly positioned: CT – maximum intensityprojection (MIP) (a) and volume rendering (b), and no endoleak was detected. After 12 months follow-up a type III endoleak (arrow) was observed: DSA (c).After EVG repositioning there was complete endoleak resolution. No contrast media extravasation was observed: CT scan (d). Successive patient’s follow-up wasuneventful.

Author's personal copy

G. Dialetto et al. / European Journal of Radiology 64 (2007) 65–72 71

experience confirm the low procedural risk of the endovasculartechnique [22–24].

The present study shows the efficacy of EVG in the long-term follow-up, with an overall survival of 82.1%, which iscomparable to that reported in recent studies [7,25].

In our series only in four patients (two in the early phase, andtwo in the late phase) the cause of death could be related to theprocedure.

We had a high mortality in the early phase (10.7%) that couldbe explained by the fact that we didn’t use restrictive enrolmentcriteria and because we performed cases in emergency condi-tions with many associated comorbidities. In fact recent studieshave demonstrated the feasibility of endovascular thoracic aor-tic repair either in emergency and in elective aortic repair as aneffective alternative to open surgical approach with a better out-come especially in the long-term follow-up [26]. Furthermorein many patients the indication to EVG treatment was put late inthe course of illness, after a prolonged phase of medical therapy[18,25].

If a patient is candidate for stent-graft treatment, three mainfactors have to be considered.

1. It is fundamental to locate either the wall disruption or intimaltear.

2. It is necessary that there is a distal vascular access of sufficientsize and a limited tortuosity of the abdominal and thoracicaorta.

3. Attention has to be paid to determination of the proximal anddistal “landing” zones of the stent-graft because they act asfriction anchors at each end.

We did not observe any case of paraplegia. At present, it isnot known which effect of the endovascular repair results inthe development of paraplegia. However, it does not seem to berelated to coverage of the intercostal arteries, whereas emboliccauses, blood loss, or insufficient collateral circulation appearto expose the patient to a major risk of spinal damage. Findingsfrom the Stanford experience and some initial reports in theliterature suggest that a 1–3% incidence of paraplegia shouldbe expected. In particular, if simultaneous open abdominal andendovascular thoracic repair is performed, the loss of lumbarand thoracic intercostal arteries increases the risk of paraplegiacaused by insufficient collateral circulation.

7. Conclusion

The evolution of endovascular techniques offers a potentialopportunity for treatment even in high-risk patients. Becauseof considerable morbidity and mortality of conventional opensurgical repair of the descending aorta, patients with TAD mightbenefit even more from an endovascular alternative than patientswith an abdominal aortic aneurysm.

When the aortic arch is involved, surgical and endovascu-lar procedures can be combined and performed simultaneously,allowing treatment of a wider range of patients. Our experience,in accordance with literature data, suggests that this techniquecan now be considered an effective alternative to open surgery.

However, long-term follow-up based on large series is necessaryto assess the durability of both the device and the repair afterstent-graft placement in the thoracic aorta. We expect that a care-ful patient selection and new, more refined stent-graft deviceswill provide better clinical results in the future.

References

[1] Ishida M, Kato N, Hirano T, Cheng SH, Shimono T, Takeda K. Endovas-cular stent-graft treatment for thoracic aortic aneurysms: short- to midtermresults. J Vasc Interv Radiol 2004;15:361–7.

[2] Czermak BV, Waldenberger P, Perkmann R, et al. Placement of endovascu-lar stent-grafts for emergency treatment of acute disease of the descendingthoracic aorta. Am J Roentgenol 2002;179:337–45.

[3] Rubin GD. MDCT imaging of the aorta and peripheral vessels. Eur J Radiol2003;45(Suppl. 1):S42–9.

[4] Bortone AS, Schena S, D’Agostino D, et al. Immediate versus delayedendovascular treatment of post-traumatic aortic pseudoaneurysms and typeB dissections: retrospective analysis and premises to the upcoming Euro-pean trial. Circulation 2002;106:34–240.

[5] Neuhauser B, Perkmann R, Greiner A, et al. Mid-term results after endovas-cular repair of the atherosclerotic descending thoracic aortic aneurysm. EurJ Vasc Endovasc Surg 2004;28:146–53.

[6] Nienaber CA, Fattori R, Lund G, et al. Non surgical reconstructionof thoracic aortic dissection by stent graft placement. N Engl J Med1999;340:1539–45.

[7] Dake MD. Endovascular stent-graft management of thoracic aortic dis-eases. Eur J Radiol 2001;39(1):42–9.

[8] Gotway MB, Dawn SK. Thoracic aortic imaging with multislice CT. RadiolClin North Am 2003;41:521–43.

[9] Therasse E, Soulez G, Giroux MF, et al. Stentgraft placement forthe treatment of thoracic aortic diseases. RadioGraphics 2005;25:157–73.

[10] Garzon G, Fernandez-Velilla M, Marti M, Acitores I, Ybanez F, Riera L.Endovascular stent-graft treatment of thoracic aortic disease. RadioGraph-ics 2005;25:s229–43.

[11] Hutschala D, Fleck T, Czerny M, et al. Endoluminal stent graft placementin patient with acute aortic dissection type B. Eur J Cardiothorac Surg2002;21:964–9.

[12] Bortone AS, Schena S, Mannatrizio G, et al. Endovascular stent-graft treat-ment for diseases of the descending thoracic aorta. Eur J Cardiothorac Surg2001;20:514–9.

[13] Dialetto G, Covino FE, Scognamiglio G, et al. Treatment of type B aor-tic dissection: endoluminal repair or conventional medical therapy? Eur JCardiothorac Surg 2005;27:826–30.

[14] White GH, Yu W, May J. Endoleak: a proposed new terminology to describeincomplete aneurysm exclusion by an endoluminal graft. J Endovasc Surg1996;3:124–5.

[15] White GH, Yu W, May J, Chaufour X, Stephen MS. Endoleak as a complica-tion of endoluminal grafting of abdominal aortic aneurysms: classification,incidence, diagnosis, and management. J Endovasc Surg 1997;4:152–68.

[16] Bortone AS, De Cillis E, D’Agostino D, de Luca Tupputi Schinosa L.Endovascular treatment of thoracic aortic disease: four years of experience.Circulation 2004;110:II262–7.

[17] Kos X, Bouchard L, Otal P, et al. Stent-graft treatment of penetratingthoracic aortic ulcers. J Endovasc Ther 2002;9(suppl 2):II25–31.

[18] Demers P, Miller DG, Mitchell RS, Kee ST, Chagonjian L, Dake MD.Stent-graft repair of penetrating atherosclerotic ulcers in the descendingthoracic aorta: mid-term results. Ann Thorac Surg 2004;77:81–6.

[19] Faries PL, Lang E, Ramdev P, Hollier LH, Marin ML, Pomposelli Jr FB.Endovascular stent-graft treatment of a ruptured thoracic aortic ulcer. JEndovasc Ther 2002;9(Suppl. 2):II20–4.

[20] Svensson LG, Crawford ES. Statistical analyses of operative results. In:Cardiovascular and vascular disease of the aorta. Philadelphia, PA: Saun-ders; 1997. pp. 432–55.

Author's personal copy

72 G. Dialetto et al. / European Journal of Radiology 64 (2007) 65–72

[21] Dake MD, Miller DC, Semba CP, Mitchell RS, Walker PJ, Liddell RP.Transluminal placement of endovascular stent-grafts for the treatment ofdescending thoracic aortic aneurysms. N Engl J Med 1994;331:1729–34.

[22] Ehrlich M, Grabenwoeger M, Cartes-Zumelzu F, et al. Endovascular stentgraft repair for aneurysms on the descending aorta. Ann Thorac Surg1998;66:19–25.

[23] Duhaylongsod FG, Glower DD, Wolfe WG. Acute traumatic aor-tic aneurysm: the Duke experience from 1970 to1990. J Vasc Surg1992;15:331–3.

[24] Grabenwoger M, Hutschala D, Ehrlich MP, et al. Thoracic aorticaneurysms: treatment with endovascular self-expandable stent grafts. AnnThorac Surg 2000;69:441–5.

[25] Patel HJ, Shillingford MS, Williams DM, et al. Survival benefit of endovas-cular descending thoracic aortic repair for the high-risk patient. Ann ThoracSurg 2007;83:1628–34.

[26] Patel HJ, Williams DM, Upchurch Jr GR, et al. Long-term results from a12-year experience with endovascular therapy for thoracic aortic disease.Ann Thorac Surg 2006;82:2147–53.