1998;66:19-24 Ann Thorac SurgDietmar Petzl, Johannes Lammer, Siegfried Thurnher, Ernst Wolner and Michael Havel

Marek Ehrlich, Martin Grabenwoeger, Fabiola Cartes-Zumelzu, Michael Grimm, Endovascular stent graft repair for aneurysms on the descending thoracic aorta

http://ats.ctsnetjournals.org/cgi/content/full/66/1/19located on the World Wide Web at:

The online version of this article, along with updated information and services, is

Print ISSN: 0003-4975; eISSN: 1552-6259. Southern Thoracic Surgical Association. Copyright © 1998 by The Society of Thoracic Surgeons.

is the official journal of The Society of Thoracic Surgeons and theThe Annals of Thoracic Surgery

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

ORIGINAL ARTICLES: CARDIOVASCULAR

Endovascular Stent Graft Repair for Aneurysms onthe Descending Thoracic AortaMarek Ehrlich, MD, Martin Grabenwoeger, MD, Fabiola Cartes-Zumelzu, MD,Michael Grimm, MD, Dietmar Petzl, MD, Johannes Lammer, MD, Siegfried Thurnher,Ernst Wolner, MD, and Michael Havel, MDDepartments of Cardio-Thoracic Surgery, Medicine, and Interventional Radiology, University of Vienna, Vienna, Austria

Background. The traditional treatment of aneurysms ofthe descending thoracic aorta includes posterolateralthoracotomy and aortic replacement with a prostheticgraft. In this study, we report our experiences and resultsin endovascular stent graft placement as an alternative tosurgical repair.

Methods. Between January 1989 and July 1997, a total of68 patients (24 women) underwent replacement of thethoracic aorta. Mean age at operation was 51 years.Fifty-eight patients underwent conventional surgicaltreatment. All of these patients were suitable candidatesfor endovascular stenting; however, no stent graft mate-rial was available at the time of operation. Ten patients (1chronic dissection, 9 atherosclerotic aneurysm) receivedin the past 8 months the first commercially manufacturedendovascular stent graft. The mean diameter of theaneurysms in this group was 7 cm (range, 6 to 8 cm). Twostent patients were operated on using only spinal cordanalgesia. All stent grafts were custom designed for eachof the 10 patients.

Results. The 30-day mortality in the conventionalgroup was 31% versus 10% in the stent group. Meanlength of intervention was 320 minutes in the conven-tional group versus 150 minutes in the endovascular

group. Spinal cord injury occurred in 5 patients (12%) inthe surgical group, whereas none of the stented patientsdeveloped any neurologic sequelae. Mean intensive careunit stay was 13 days, followed by a mean of 10 days ona ward in the first group compared to 4 days in theintensive care unit and 6 days on the ward in the stentgroup. One stent was required in 2 patients, two stentswere required in 3 patients, and four stents were de-ployed in 5 patients of our series. Five patients requiredtransposition of the left subclavian artery to achieve asufficient neck for the proximal placement of the stent.There was complete thrombosis of the thoracic aorticaneurysm surrounding the stent graft in 8 patients (80%).Two patients required restenting as a result of leakage(20%). Stent graft placing was performed through thefemoral artery in 8 patients, whereas access was onlyachieved through the abdominal aorta in 2 patients.

Conclusions. These preliminary results demonstratethat endovascular stent graft replacement might be apromising, cheaper, and safe alternative method in se-lected patients with descending thoracic aneurysms.

(Ann Thorac Surg 1998;66:19–25)© 1998 by The Society of Thoracic Surgeons

Descending thoracic aortic aneurysms still represent apotentially life-threatening situation. Resection

and graft replacement of the pathologically altered aortais the preferred method of treatment [1–6]. Althoughgreat strides have been achieved during the past decadesin the management of patients with thoracic aortic aneu-rysms by new surgical techniques, intraoperative moni-toring of somatosensory-evoked potentials, and spinalcord fluid drainage, postoperative morbidity and mortal-ity rates still remain high [7–10]. The afflicted populationis usually of older age and present at time of operationwith various comorbidities such as hypertension, ob-structive pulmonary disease, coronary heart disease, allof which have significant impact on the surgical outcome.Postoperative complications, such as paraplegia, renal

and pulmonary insufficiency, contribute to prolongedhospital stays and higher medical cost.

Since the first successful endovascular stent graftplacement in an abdominal aortic aneurysm [11], variousgroups have started to investigate the feasibility of tho-racic aortic aneurysmal repair with transluminally placedstent grafts [12, 13]. This study was undertaken to reportour initial clinical experience with the repair of descend-ing thoracic aortic aneurysms with a commercially avail-able endovascular stent graft and compare these resultswith a conventionally treated group.

Material and Methods

PatientsBetween January 1989 and July 1997, a total of 68 patientsunderwent replacement of the descending thoracic aortaat the University of Vienna Medical Center. Patients weregrouped into two categories based on perioperative man-agement. Group 1 consisted of all patients operated on by

Presented at the Thirty-fourth Annual Meeting of The Society of ThoracicSurgeons, New Orleans, LA, Jan 26–28, 1998.

Address reprint requests to Dr Ehrlich, Department of Cardio-ThoracicSurgery, Mount Sinai Medical Center, One Gustave Levy Place, Box 1028,New York, NY 10029 (e-mail: mehrl98000@aol.com).

© 1998 by The Society of Thoracic Surgeons 0003-4975/98/$19.00Published by Elsevier Science Inc PII S0003-4975(98)00390-7

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

conventional surgical treatment. All of the 58 patients inthis group were suitable candidates for endovascularstenting; however, no stent graft material was available atthe time of operation. Group 2 comprised 10 patients whoreceived in the past 8 months a commercially manufac-tured endovascular stent-graft. Table 1 summarizes im-portant preoperative characteristics of both groups. Hy-pertension was the most common preoperative medicaldisorder with an incidence of 73% in all patients, fol-lowed by a history of smoking (66%) and coronary heartdisease (33%).

The conventional technique used in the first groupincluded profound hypothermic circulatory arrest in 42patients, left heart bypass in 5 patients, and femoral/femoral nonheparinized roller pump in 11 patients, re-spectively.

All stent graft patients gave their informed consent inconformance with the protocols approved by the institu-tional review board of Vienna Hospital. They underwentchest radiography, spiral computed tomographic scan-ning, and arteriography before the procedure and atintervals after the procedure. Patients eligible for receiv-ing a stent graft had to have at least a 1- to 2-cm longproximal and distal neck. Chronic aortic dissection wasnot an exclusion criteria in our stent series; however, thisprocedure was only performed when the false lumen wasalready completely thrombosed. Acute aortic dissectionwas a contraindication for stent graft deployment.

The mean stent graft diameter was 36 mm (range, 32 to40 mm), and the mean length of the device was 104 mm(range, 100 to 120 mm). The preoperative characteristicsof the 10 stented patients are shown in Table 2.

Endovascular Grafting SystemThe endovascular stent graft (Talent Endoluminal SpringGraft System, World Medical Corporation, Sunrise, FL)was custom designed for each patient (Fig 1). It is aself-expanding spring-stent graft that is compressed overa multiple-lumen polyurethane placement catheter. Thestent graft and catheter are loaded into a hollow sheath.The entire system is introduced into the vasculature andadvanced to the aneurysmic site. The system consists of aplacement catheter that has one through lumen used tocarry a guidewire. The catheter also contains a polyure-thane tip and central balloons. The central balloon isused to model the spring stent graft to the vessel wallafter deployment. The graft by itself is made of polyester(Dacron) vascular graft and a nitinol wire stent, withspring shaped in a zig-zag serpentine formation andplaced approximately 5 to 15 mm apart. A straight nitinolconnecting wire extends from the proximal-most springto the distal-most spring and serves to fix the length ofthe device and avoid twisting and kinking. Furthermore,the system consists of a hollow introducer sheath. Thesheath includes a hemostasis valve and a side port withstopcock. The sheath’s stiffness can be controlled byinjecting/withdrawing a solution into/from the side port.Finally, the system contains a push rod with plunger thatincludes a Touhy-Borst valve and a plunger. The valveholds the plunger in place and prevents excess bloodleakage.

Fig 1. Stent graft material used in our series.

Table 1. Selected Preoperative Clinical Characteristics andComplications

CharacteristicConventional

GroupStent

Group

Age (y) 62 6 16 68 6 11Male:female (n) 37:21 7:3COPD 19% 20%History of smoking (%) 67% 60%Hypertension 72% 80%Renal insufficiency 19% 20%Coronary heart disease 24% 40%Previous myocardial infarction 29% 60%

COPD 5 chronic obstructive pulmonary disease.

Table 2. Characteristics of all 10 Patients Who Received Endovascular Stent Graft Placement

PatientAge(y) Sex

Type ofAneurysm

Preoperative Diameterof Aneurysm (mm)

Mean StentDiameter (mm)

Mean StentLength (mm)

1 69 M True 60 36 1102 76 M True 60 37 1053 61 M True 70 37 102.54 70 F True 60 36.5 107.55 70 M True 70 34 1006 40 M True 80 38 1057 63 M Dissection 70 37.5 102.58 77 F True 80 37 1209 76 F True 70 37 110

10 71 M True 80 32 110

20 EHRLICH ET AL Ann Thorac SurgSTENT GRAFT FOR THORACIC AORTIC ANEURYSM 1998;66:19–25

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

Endovascular ProcedureAll but 2 patients were under general anesthesia, andcardiopulmonary bypass standby was available in everycase. The largest femoral artery is exposed and controlledproximally and distally. If the femoral artery is consid-ered too small in diameter, the iliac artery or abdominalaorta is used for vascular access. Five thousand units ofheparin is injected intravenously. Using a percutaneousleft brachial artery approach, a catheter is inserted over aguidewire localizing the subclavian artery and celiac axisand is used for aortogram. Through a transverse femoralarteriotomy, a guidewire is inserted and passed abovethe aneurysmal site. With use of image amplification, theendovascular stent system is passed over the guidewireand positioned at the desired location. The guidewire isremoved. An arteriography is performed to verify theexact location of the proximal stent. By holding the pushrod and placement catheter stationary and inflating thetip balloon, the sheath is moved downward to deploy theproximal spring. The nitinol spring expands and con-forms to the shape and size of the proximal neck of theaortic aneurysm and will maintain sufficient pressureagainst the aorta to create and sustain a blood-tight seal.The balloon tip is slowly deflated (Fig 2). With gentledownward movement of the placement catheter, thegraft is opened. The integrated balloons are used toocclude blood flow to the deployment site, fix the posi-tion of the proximal stent during deployment, model thespring stent to the vessel wall neck, and smooth wrinklesin the graft material. The push rod and placement cath-eter are held stationary and the sheath is moved down-ward. The distal nitinol spring is deployed and balloonedto seal against the distal aortic aneurysm neck. A secondarteriography is performed to verify sealing of the graft.The placement system is removed, the arteriotomy isclosed in the usual manner. Figure 3 shows graphically a

typical stent graft at the end of the procedure. Beforedischarge, aortogram and spiral computer tomography isperformed to document aneurysmal exclusion.

StatisticsData are presented as percentage or means 6 standarddeviation. The data from the two groups were comparedwith unpaired Student’s t test. All differences were re-garded as statistically significant if p was less than 0.05.

Results

The 30-day mortality in the conventional group was 31%versus 10% in the stent group. Table 3 summarizesimportant postoperative parameters between bothgroups. Patients treated conventionally by left thoracot-omy had longer intensive care unit as well as hospitalstays. The duration of the procedure was much shorter inthe stented group and there were no neurologic compli-cations associated with endovascular stent graft place-ment. However, all of our 10 patients in the stent grouphad proximal descending aneurysms where covering ofand as a consequence sacrifice of intercostal arteries withthe stent graft is not of so much importance as in distaldescending thoracic aortic repair. Three of the 5 patientswith spinal complications in the conventional group hadextension of the aneurysm into the distal descendingaorta with reattachment of intercostal arteries at a level ofT-11 and T-12. Four of these 5 patients have been oper-ated with the use of profound hypothermic circulatoryarrest and 1 patient had been operated with left heartbypass.

The deployment of the stent graft at the intendedposition was successfully performed in all 10 patients.Typical aortograms obtained before and after the proce-dure are shown in Figures 4 and 5. Stent graft placing was

Fig 2. Proximal stent graft deployment at the proximal descendingaorta.

Fig 3. Typical stent graft after final accomplishment of the proce-dure.

21Ann Thorac Surg EHRLICH ET AL1998;66:19–25 STENT GRAFT FOR THORACIC AORTIC ANEURYSM

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

carried out through the femoral artery in 4 patients, iliacartery in 4 patients, whereas access was only achievedthrough the abdominal aorta in 2 patients (Table 4).Introduction of the sheath caused dissection of the iliacartery in 1 patient (patient 9). The sheath was removedimmediately and a iliac–femoral bypass was performed.Access to the descending aorta was accomplished in thesame patient from a retroperitoneal approach with intro-duction of the sheath directly into the abdominal aorta.Complete aneurysmal thrombosis was accomplished im-mediately in 8 patients (80%). Two patients had leakageinto the aneurysmal sac. Patient 5 had a small proximaltract communicating with the aneurysm at the distalaortic arch. Patient 8 had leakage into the aneurysmal sacbetween two stents. Both patients required a secondprocedure and a separate stent graft prosthesis wasneeded to cover the residual filling site.

There was one postoperative death in the stentedgroup. This was in a 76-year-old man with an atheroscle-rotic aneurysm 60 mm in diameter (patient 2). No trans-position of the subclavian artery had to be carried outand the patient received two stent grafts introduced

through the femoral artery. He had an uneventful post-operative recovery in the intensive care unit and wastransferred to the ward on the second day after theprocedure. The same patient had rupture of the descend-ing thoracic aorta on postoperative day 7 as a result ofstent penetration through the aneurysmatic aortic walland died immediately after the event. As of this writing,none of the 9 surviving stent patients have shown aserious complication such as graft migration, aneurysmrupture, aneurysms enlargement, or a change in theposition and configuration of the graft.

Comment

Since 1969, when Dotter [14] inserted stainless steel coilsas a vascular stent in canine popliteal arteries, translu-minal placement of endovascular stent grafts has beenused as a treatment of abdominal aortic aneurysms aswell as recently for descending thoracic aneurysms ordissections [12, 15]. Although long-term follow-up islimited at present, early experimental as well as clinicalresults have been promising [16, 17]. The aim of thisminimally invasive technique is to achieve aneurysmalexclusion, to suppress the pressure stress on the aneu-rysmal sac, and to reduce mortality and morbidity ratescompared with those associated with standard surgicaltreatment. Stent graft placement might be of particularbenefit, especially in high-risk patients with descendingthoracic aneurysms, because this procedure is easier andless invasive than the conventional surgical approach. Itenables avoidance of left thoracotomy with the conse-quences of lung manipulation and compression as wellas phrenic and recurrent nerve injury resulting in pro-longed postoperative respiratory support. Coagulopa-thies associated with deep hypothermia and circulatoryarrest can also be avoided by this technique.

Fig 4. Preoperative aortogram from a patient with a descending tho-racic and left subclavian aneurysm.

Fig 5. Postoperative aortogram from the same patient after transpo-sition of the left subclavian artery and stent graft placement.

Table 3. Intraoperative and Postoperative ParametersBetween Both Groupsa

Parameter Conventional Stent p Value

Length of intervention (min) 320 6 94 150 6 28 ,0.05Mean length of intensive care

unit stay (days)13 6 15 4 6 2 ,0.05

Mean hospital stay (days) 10 6 3 6 6 1 ,0.05Spinal cord injury (%) 12 0 NSOperative mortality (%) 31 10 NS

a Values are mean 6 standard deviation.

NS 5 not significant.

22 EHRLICH ET AL Ann Thorac SurgSTENT GRAFT FOR THORACIC AORTIC ANEURYSM 1998;66:19–25

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

If a patient is considered to be a candidate for endo-vascular stent graft placement, a number of anatomic aswell as technical matters have to be taken into account. Itis imperative to perform preoperative imaging studies togain information about the precise vascular access andaneurysmal size. Longitudinal and transverse diameterof the aneurysm provide information that are needed todesign a precise stent prosthesis for each individualpatient. Furthermore, an aortogram should be performedbefore the procedure to determine the anatomic relation-ship to major aortic branches, such as the left subclavianartery or the celiac axis.

The perioperative mortality rate could be decreased inthis series from 31% in the conventional-treated group to10% in the endovascular stent graft group, but did notreach statistical significance owing to the small numberof patients in the latter group. These results are similar toa recent stent study carried out by Mitchell and col-leagues [12] from Stanford where the mortality rate in 44patients was 7%. Furthermore, the majority of our pa-tients who received an endovascular stent graft wereconsidered poor candidates for conventional open repairbecause of various comorbidities such as previous myo-cardial infarction, hypertension, and coronary heart dis-ease.

The incidence of paraplegia in descending thoracicaneurysms is reported from 0 to 18% [10, 18, 19]. Therewas no instance of paraplegia in our stent-treated pa-tients. Although there is no opportunity to reimplantintercostal arteries with this new technique, most of theseintercostal branches are usually thrombosed in athero-sclerotic aneurysms at the time of intervention. It isunknown at this point which effect endoluminal graftingwill have on spinal cord injury. However, aortic cross-clamping for a longer period with the consequences ofspinal cord hypoperfusion can be avoided by this newprocedure.

At this time, there are still pitfalls that have to beaddressed with this surgical approach. The first concernsthe device itself. In its present state, graft deployment isstill relatively inexact and precise stent placement issometimes hard to achieve. Miniaturization of the stent

device has to be developed to prevent complications suchas vascular trauma during introduction.

Another problem concerns perigraft leakage. Thiscomplication occurred in our study in 2 patients. Themain cause of this problem was the limited anatomicsuitability in 1 patient, which made it difficult to achievea precise placement of the stent graft as well as distalmigration of one stent after deployment in the secondpatient. Although it remains uncertain at this stage ofknowledge whether leakage into the aneurysmal sackmay lead to late aortic rupture, both patients with stentgraft leakage had a second procedure right after the firstintervention to obliterate the leakage spot.

It is our belief that an absolute contraindication forendovascular stent graft placement is a severe tortuosityof the descending thoracic aorta or iliac arteries whereintroduction of the stent device to the desired locationmight be associated with a risk of aortic or arterialrupture. Furthermore, patients with acute aortic dissec-tions were excluded for stent graft placement in ourstudy. Although 1 patient with a chronic dissection,where the false lumen was already thrombosed, receiveda stent graft, it remains uncertain whether this techniqueis suitable for this specific patient cohort. Furthermore,should this procedure be limited to patients with asufficient distal and proximal neck and the absence ofnearby important side branches. Therefore, transpositionof the left subclavian artery was performed in 5 patients.

Two patients, where stent graft insertion was per-formed through the femoral artery, were operated on bythe use of spinal cord analgesia. It is our belief, however,that this anesthetic technique should be limited to thosepatients, where access into the vascular system isplanned from the groin. Manipulation of the iliac arteryor abdominal aorta for graft insertion from a retroperito-neal approach might be associated with severe pain andshould be performed only by general anesthesia.

Although our results with stent graft placement on thedescending thoracic aorta demonstrated that this methodis feasible, there remain questions and long-term consid-erations. Short- and long-term biologic interactions be-

Table 4. Characteristics of all 10 Patients Who Received Endovascular Stent Graft Placement

PatientNo. Anesthesia

Location ofAccess

Transposition ofSubclavian Artery Complication

1 GA Iliac artery No None2 SCA Femoral artery No Rupture3 GA Iliac artery Yes None4 GA Abdominal aorta No None5 GA Iliac artery Yes Leak6 SCA Femoral artery Yes None7 GA Femoral artery Yes None8 GA Abdominal aorta Yes Leak9 GA Femoral artery No Dissection of iliac artery

10 GA Iliac artery No None

GA 5 general anesthesia; SCA 5 spinal cord anesthesia.

23Ann Thorac Surg EHRLICH ET AL1998;66:19–25 STENT GRAFT FOR THORACIC AORTIC ANEURYSM

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

tween prosthesis and aortic wall need further investiga-tion.

In summary, our results demonstrate that insertion ofan endovascular stent graft into the descending aorta isfeasible, and that intraluminal aneurysm exclusion canbe achieved acutely. Further studies need to be per-formed to determine the long-term effectiveness. Thistechnique may become applicable for treating descend-ing aortic aneurysm, especially in high-risk patients, asthis approach is less invasive than the current conven-tional surgical methods.

References

1. Cooley DA, DeBakey ME. Surgical considerations of in-trathoracic aneurysms of the aorta and great vessels. AnnSurg 1952;135:660–80.

2. DeBakey ME, Cooley DA. Successful resection of aneurysmof the thoracic aorta and replacement by graft. JAMA 1953;152:673–6.

3. Cooley DA, DeBakey ME. Surgical consideration of exci-sional therapy for aortic aneurysms. Surgery 1953;34:1005–20.

4. Livesay JJ, Cooley DA, Ventemiglia R, et al. Surgical experi-ence in descending thoracic aneurysmectomy with and with-out adjuncts to avoid ischemia. Ann Thorac Surg 1985;39:37–46.

5. Crawford ES, Walker HSJ, Saleh SA, Normann NA. Graftreplacement of aneurysm in descending thoracic aorta: re-sults without bypass or shunting. Surgery 1981;89:73–85.

6. Najafi H, Javid H, Hunter J, Serry C, Monson D. Descendingaortic aneurysmectomy without adjuncts to avoid ischemia.Ann Thorac Surg 1980;30:326–35.

7. Hollier LH, Money SR, Naslund TC, et al. Risk of spinal corddysfunction in patients undergoing thoracoabdominal aorticreplacement. Am J Surg 1992;164:210–4.

8. Crawford ES, Mizrahi EM, Hess KR, Coselli JS, Safi HJ, Patel

VM. The impact of distal aortic perfusion and somatosen-sory evoked potential monitoring on prevention of paraple-gia after aortic aneurysm operation. J Thorac CardiovascSurg 1988;95:357–67.

9. Kouchoukos NT, Wareing TH, Izumoto H, Klausing W,Abboud N. Elective hypothermic cardiopulmonary bypassand circulatory arrest for spinal cord protection duringoperations on the thoracoabdominal aorta. J Thorac Cardio-vasc Surg 1990;99:659–64.

10. Borst HG, Jurmann M, Buhner B, Laas J. Risk of replacementof descending aorta with a standardized left heart bypasstechnique. J Thorac Cardiovasc Surg 1994;107:126–33.

11. Parodi JC, Palmaz JC, Barone HD. Transfemoral intralumi-nal graft implantation for abdominal aortic aneurysms. AnnVasc Surg 191;5:491–9.

12. Mitchell RS, Dake MD, Semba CP, et al. Endovascularstent-graft repair of thoracic aortic aneurysms. J ThoracCardiovasc Surg 1996;111:1054–62.

13. Inoue KI, Iwase T, Sato M, et al. Clinical application oftransluminal endovascular graft placement for aortic aneu-rysms. Ann Thorac Surg 1997;63:522–8.

14. Dotter CT. Transluminally-placed coilspring endoarterialtube grafts: long-term patency in canine popliteal artery.Invest Radiol 1969;4:329–32.

15. Dake MD, Miller DC, Semba CP, Mitchell RS, Walker PJ,Liddell RP. Transluminal placement of endovascular stent-grafts for the treatment of descending thoracic aortic aneu-rysms. N Engl J Med 1994;331:329–34.

16. Moon MR, Dake MD, Pelc LR, et al. Intravascular stenting ofacute experimental type B dissections. J Surg Res 1993;54:381–8.

17. Fann JI, Dake MD, Semba CP, Liddell RP, Pfeffer TA, MillerDC. Endovascular stent-grafting after arch aneurysm repairusing the “elephant trunk”. Ann Thorac Surg 1995;60:1102–5.

18. DeBakey ME, McCollum CH, Graham JM. Surgical treat-ment of aneurysms of the descending aorta. J CardiovascSurg (Torino) 1978;19:571–6.

19. Moreno-Cabral CE, Miller DC, Mitchell RS, et al. Degener-ative and atherosclerotic aneurysms of the thoracic aorta.J Thoracic Cardiovasc Surg 1984;88:1020–32.

DISCUSSION

DR D. CRAIG MILLER (Stanford, CA): It is a pleasure to beasked to discuss Dr Ehrlich’s paper. For those of you who do notknow him, Dr Ehrlich is a third-year surgical resident in ErnstWolner’s group at the University of Vienna, and I complimenthim on a fine presentation.

Because outside of our center and Japan almost nobody elsehas explored endovascular stent grafting for descending thoracicaneurysms, it is refreshing to hear the Austrian experience usinga Talent device. This is the only way we will ever figure outwhether this approach has merit, and, if so, in which patients.The Vienna experience is small but does confirm our findings in122 patients treated similarly with a stent graft between 1992 and1997.

Stent grafting of the descending thoracic aorta is feasible andsafe. Because of anatomic features, however, it is not advisablefor all patients, which leads me to my first question. How couldall 58 earlier patients in your series going back to 1989 treatedwith open surgical techniques have been “suitable candidatesfor a stent graft” on anatomic grounds?

Second, and in a similar vein, in the 8 months since you havebeen doing stent grafts, how many other patients required opensurgical repair? This is not for everybody, as I am sure some ofthe audience knows.

Third, the Vienna group early in their experience was quiteaggressive in creating a proximal neck or a “stent graft landing

zone” by transposing the left subclavian artery in 50% ofpatients. In our experience we have resorted to this option inonly about 12% of patients where there just is not enough neckbeyond the left subclavian artery for a safe landing zone.Nevertheless, this indicates to me that most of your aneurysmsinvolved the very proximal descending thoracic aorta, such thatspinal cord injury would be expected to be very rare. Therefore,observing no cases of paraplegia in your initial 10 patients is notsurprising. On the other hand, in our experience where occa-sionally the entire descending thoracic aorta has been covered,the paraparesis/paraplegia rate is 3%. Interestingly, all 4 of theseunfortunate patients underwent simultaneous or previous ab-dominal aortic aneurysm repair. The audience should note thatstent grafting does not confer any lower risk of spinal cord injurythan does open surgical graft replacement, as Dr Ehrlich’scomparisons might be interpreted.

Fourth, your primary success and secondary success rateswere 80% and 90%, with 2 patients requiring additional stentgrafts for early endoleaks. These figures parallel our numbers of83% when leaving the operating room and 96% at hospitaldischarge. Ten of our patients required coil embolization forsmall endoleaks, and 6, another stent graft. But this leads me toan important point. In the article you state, “It is uncertain ifresidual leakage into the aneurysm sac may lead to late aorticrupture.” I am afraid we learned the hard way that it is

24 EHRLICH ET AL Ann Thorac SurgSTENT GRAFT FOR THORACIC AORTIC ANEURYSM 1998;66:19–25

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from

1998;66:19-24 Ann Thorac SurgDietmar Petzl, Johannes Lammer, Siegfried Thurnher, Ernst Wolner and Michael Havel

Marek Ehrlich, Martin Grabenwoeger, Fabiola Cartes-Zumelzu, Michael Grimm, Endovascular stent graft repair for aneurysms on the descending thoracic aorta

& ServicesUpdated Information

http://ats.ctsnetjournals.org/cgi/content/full/66/1/19including high-resolution figures, can be found at:

References http://ats.ctsnetjournals.org/cgi/content/full/66/1/19#BIBL

This article cites 18 articles, 9 of which you can access for free at:

Citations http://ats.ctsnetjournals.org/cgi/content/full/66/1/19#otherarticles

This article has been cited by 31 HighWire-hosted articles:

Permissions & Licensing

.healthpermissions@elsevier.comemail: orhttp://www.us.elsevierhealth.com/Licensing/permissions.jsp

in its entirety should be submitted to: Requests about reproducing this article in parts (figures, tables) or

Reprints reprints@elsevier.com

For information about ordering reprints, please email:

by on June 2, 2013 ats.ctsnetjournals.orgDownloaded from