Reintervention after thoracic endovascular aortic

repair of complicated aortic dissection

Elsa M. Faure, MD,a,b Ludovic Canaud, MD, PhD,a,b Camille Agostini, MD,c Roxane Shaub, MD,c
Gudrun Böge, MD,a Charles Marty-ané, MD, PhD,a and Pierre Alric, MD, PhD,a,b Montpellier, France

Objective: This study assessed predictive factors for reintervention after thoracic endovascular aortic repair (TEVAR) for
complicated aortic dissection (C-AD).
Methods: An institutional review of consecutive TEVAR for C-AD was performed.
Results: Between 2000 and 2011, 41 patients underwent TEVAR for a C-AD involving the descending thoracic aorta.
Primary indications included aneurysm >55 mm in 24, rapid aneurysmal enlargement or impending rupture in 6, saccular
aneurysm >20 mm in 1, malperfusion in 1, intractable chest pain in 3, and rupture in 6. Technical success was achieved in
100%. The 30-day mortality rate was 5% (n [ 2). Fourteen secondary procedures were performed in 13 patients (32%) for
indications of device migration in 2, proximal type I endoleak in 5, distal type I endoleak in 2, type II endoleak in 1,
aneurysmal evolution of the descending thoracic aorta in 2, aneurysmal expansion of the dissected abdominal aorta in 1,
and retrograde dissection in 1. Multivariate analysis demonstrated that oversizing $20% (odds ratio [OR], 16; P [ .011),
bare-spring stent in the proximal landing zone of the stent graft (OR, 12; P [ .032), and anticoagulant therapy (OR, 78;
P [ .03) were significant factors for reintervention. On univariate analysis, large aneurysm was a risk factor for rein-
tervention (P [ .002), whereas complete false lumen thrombosis at the stent graft level was protective (P < .05).
Conclusions: This study confirms the feasibility of TEVAR for C-AD, although the rate of reintervention is high. Excessive
oversizing, a bare-spring stent graft in the proximal landing zone, large aortic dilatation, and anticoagulant therapy were
factors associated with reintervention. Complete false lumen thrombosis at the stent graft level was protective. (J Vasc
Surg 2014;59:327-33.)

Initially designed to treat degenerative aneurysms, compared with those without ($3668) secondary proce-
thoracic endovascular aortic repair (TEVAR) was intro- dures.8 It is important that significant clinical and technical
duced as an alternative minimally invasive procedure for factors associated with secondary intervention are identified
the treatment of complicated aortic dissection (C-AD) in and mechanisms of failure elucidated so that we may
1999.1 Endovascular repair works by covering the proximal prevent and manage them in future. Published data in
tear, thus reducing or preventing flow in the false lumen this area have been lacking.
and thereby allowing true lumen expansion. As a conse- The aim of our study was to identify the risk factors for
quence of lower perioperative morbidity and mortality reintervention after TEVAR for the treatment of C-AD
compared with open surgery, consensus has now shifted involving the descending thoracic aorta.
such that many now consider TEVAR as the first-line
therapy for C-AD.2,3 METHODS
However, the long-term durability of this endovascular
The Institutional Review Board approved this study,
approach is still being debated. Some reports suggest a high
and informed consent was obtained from each participant.
rate of reintervention4-6 of up to 46% in the acute setting.7
Patients selection and follow-up. The study included
This has raised concerns regarding the long-term outcomes
all consecutive patients undergoing TEVAR for C-AD,
and costs of TEVAR for C-AD. Recent cost-effectiveness
defined as aortic aneurysm >55 mm, rapid aneurysmal
analyses have demonstrated that the postintervention costs
enlargement or impending rupture (>5 mm over 6 months),
of TEVAR are increased eightfold in those with ($31,696)
rupture, saccular aneurysm >20 mm, malperfusion
syndrome, or intractable chest pain under medical therapy
From the Department of Thoracic and Vascular Surgery, University Hospi-
involving the descending thoracic aorta between October
tala; U1046, Institut National de la Santé et de la Recherche Médicale, 2000 and December 2011 at our unit.
Université Montpellier 1b; and Clinical Research and Epidemiology Aortic dissection was defined as chronic 14 days after
Department, University Hospital.c onset of acute symptoms. According to the Reporting
Author conflict of interest: none.
Standards For Thoracic Endovascular Aortic Repair,9 tech-
Reprint requests: Elsa M. Faure, MD, Service de Chirurgie Thoracique et
Vasculaire, Hôpital A de Villeneuve, 191 av du Doyen Giraud, 34295 nical success was defined as complete coverage of the
Montpellier, France (e-mail: elsafaure@hotmail.com). primary entry tear without a type I leak at the end of the
The editors and reviewers of this article have no relevant financial relationships procedure.
to disclose per the JVS policy that requires reviewers to decline review of any Patients were preoperatively evaluated with respect to
manuscript for which they may have a conflict of interest.
0741-5214/$36.00
age, sex, smoking, obesity, hypertension, diabetes, long-
Copyright Ó 2014 by the Society for Vascular Surgery. term anticoagulant therapy, renal insufficiency, history of
http://dx.doi.org/10.1016/j.jvs.2013.08.089 ascending or abdominal aortic repair, and according to

327
 JOURNAL OF VASCULAR SURGERY
328 Faure et al February 2014

the American Society of Anesthesiologists Physical Status Table I. Clinical and demographic characteristics of the
Classification. They were prospectively monitored by sched- 41 patients at the initial procedure
uled clinical observation and with contrast-enhanced
computed tomography (CT) preoperatively, before hospital Characteristics a All patients (N ¼ 41)
discharge, at 1, 3, 6, and 12 months, and annually there-
Age, years 66 6 11
after. All aortic CT measurements were taken in a perpendic- Male sex 34 (83)
ular plane using centerline reconstructions. Preoperative Hypertension 31 (76)
and postoperative measures included neck diameter and Diabetes 2 (5)
length for device sizing, maximal aortic diameter, and aortic Smoker 18 (44)
Coronary artery disease 6 (15)
diameters at the levels of the proximal thoracic aorta (2 cm Dyslipidemia 5 (12)
below the left subclavian artery ostium), the midthoracic COPD 9 (22)
aorta (at the level of the left inferior pulmonary vein), Renal insufficiency 2 (5)
the celiac artery, and both renal arteries. Extension of the Obesity 6 (15)
dissection above or below the celiac axis and false lumen Prior aortic surgery 16 (39)
Corticosteroid therapy 5 (12)
status at the level of the stent and more distally were also Anticoagulant therapy 5 (12)
analyzed. The number of visceral and renal vessels perfused Marfan syndrome 1 (2)
by the false lumen was also reported, but we did not analyze ASA classification
that variable in terms of reintervention. 2 16 (39)
3 22 (54)
With regards to outcomes, patients were grouped by
4 3 (7)
those with and without reintervention. Reintervention
was defined as any intervention after the original procedure ASA, American Society of Anesthesiologists; COPD, Chronic obstructive
that was related to the dissection or a complication of the pulmonary disease.
a
Continuous data are shown as mean 6 standard deviation and categoric
original procedure. data as number (%).
Endovascular device and procedural details. All
endovascular procedures were performed in an operating
theater under general anesthesia. Open femoral cutdowns thrombosis was considered as a variable of follow-up and
were used to facilitate catheter-based access to the true was not analyzed in the multivariate logistic regression. This
lumen and the aortic arch. Transesophageal echocardiog- variable was analyzed separately. The multivariate model
raphy was frequently used to verify the position of the was built by a step-down procedure. The adjusted odds
guidewire in the true lumen. Angiograms were performed ratios (ORs) and 95% confidence intervals (CIs) were
before stent graft deployment to clearly delineate the calculated. Univariate and multivariate logistic regressions
proximal tear site and after deployment to confirm entry were performed with the Firth penalized likelihood method
tear sealing and absence of endoleak. For patients with because some models had quasi-complete separation of data
a short proximal landing zone (<20 mm), hybrid opera- and the results of logistic regression with maximum likeli-
tions, combining extra-anatomic debranching of the supra- hood estimates are unreliable for small samples. A P value
aortic vessels with immediate endovascular stent graft of <.05 was considered statistically significant. All statistical
deployment, were performed to enable coverage of the analyses were performed at the Clinical Research and
proximal entry tear. Epidemiology Department of the Centre Hospitalier Uni-
During the period of study inclusion, five different versitaire Montpellier with SAS 9 software (SAS Institute
stent grafts were implanted, comprising the Thoracic Inc, Cary, NC).
Excluder and C-TAG (W.L. Gore and Associates, Flagstaff,
Ariz), Talent and Valiant (Medtronic, Santa Rosa, Calif), RESULTS
and Zenith TX2 (Cook, Bloomington, Ind). Stent graft Patient demographics. Between October 2000 and
selection was at the discretion of the surgeon. For the anal- December 2011, 41 patients underwent stent graft place-
ysis, the stent grafts were grouped depending on the pres- ment for a C-AD involving the descending thoracic aorta.
ence of a proximal bare spring: The Excluder, the C-TAG, Patients characteristics are reported in Table I. The patients
and the TX2 were classified as “membrane-covered prox- were a median age of 66 years (range, 42-94 years), with
imal-spring,” and the Valiant and the Talent as “proximal a male-to-female ratio of 4.9:1. Dissection characteristics
bare-spring” stent grafts. are reported in Table II. The median interval from the
Statistical analysis. All primary procedural, reinter- initial dissection to stent graft repair was 1.5 months
vention, and outcomes data were prospectively collected (interquartile range, 0.3-36 months). The dissection was
in a vascular registry. Patients who did and did not undergo treated at the acute phase in 15 patients and at the chronic
reinterventions were compared. Univariate and multivar- phase in 26. The indication for repair included aneurysm
iate logistic regressions were performed to assess the influ- >55 mm in 24, rapid aneurysmal enlargement or
ence of initial variables on the occurrence of reintervention. impending rupture (>5 mm over 6 months) in 6, saccular
Variables of initial characteristics significantly associated with aneurysm >20 mm in 1, malperfusion syndrome in 1,
reintervention by univariate logistic regression (P < .05) intractable chest pain under medical therapy in 3, and
were entered into the model. Postoperative false lumen rupture in 6.
JOURNAL OF VASCULAR SURGERY
Volume 59, Number 2 Faure et al 329

Table II. Morphologic characteristics of the dissection Table III. Details of the initial procedure of the 41
of the 41 patients at the initial procedure patients

Characteristics a All patients (N ¼ 41) Initial intervention a All patients (N ¼ 41)

Acute 15 (37) Left subclavian artery coverage 18 (44)

Aneurysm >55 mm 2 With debranching of the SAV 16 (39)
Rapid aneurysmal enlargement 3 Proximal oversizing $20% 11 (27)
Malperfusion syndrome 1 Stent graft with proximal bare spring 9 (22)
Intractable chest pain 3 Length of aortic coverage, cm 18 6 7
Rupture 6
Chronic 26 (63) SAV, Supra-aortic vessels.
Aneurysm >55 mm 22
a
Continuous data are shown as mean 6 standard deviation and categoric
Saccular aneurysm >20 mm 1 data as number (%).
Rapid aneurysmal enlargement 3
Proximal entry tear
Zone 2 10 (24) proximal landing zone (<20 mm) for a total endovascular
Zone 3 31 (76) solution. Also performed were 12 partial-arch debranch-
Extensive dissectionb 35 (85) ing procedures through cervical access and four total
Maximal aortic diameter, mm 57 6 17
Proximal landing zone arch debranching procedures with ascending aorta-
Length, mm 27 6 9 based bypass grafts to all of the arch branches through
Diameter, mm 30.7 6 5.5 sternotomy.
a
Continuous data are shown as mean 6 standard deviation and categoric
Intraoperative complication in four patients (10%)
data as number (%). included, in one patient each:
b
Extending into the abdominal aorta vs limited dissection, defined as
restricted to the thoracic aorta. d Rupture of the right external iliac artery treated with
successful deployment of a covered stent (Wallgraft
10-70);
Mean maximal aortic diameter was 57 6 17 mm. The d Device misplacement and inadequate seal of the
proximal entry tear was in zone 2 in 10 (24%) and in zone primary entry tear with an associated proximal type I
3 in 31 (76%). The dissection extended to the abdominal endoleak on aortogram due to a tortuous seal zone
aorta in 35 (85%) and was restricted to the thoracic aorta anatomy, despite a previous total debranching of the
in 6 (15%). The branch vessels arising from the false lumen supra-aortic trunks and required the successful deploy-
were the left renal artery in 13 patients (32%), the right ment of an additional proximal stent graft;
renal artery in 7 (17%), the celiac trunk in 7 (17%), and d Migration of the stent graft 5 cm caudally that was
the superior mesenteric artery in 8 (20%). successfully treated by additional proximal stent graft
Prior aortic surgery in 14 patients including 10 with placement; and
prior type A dissection or arch repair and two each with d Unintentional partial endograft coverage of the left
open abdominal aortic aneurysm repair and aortobifemoral common carotid artery successfully treated by stenting
bypass. the origin of the left common carotid artery.
Procedural outcomes. Details of the initial procedure
of the 41 patients are reported in Table III. The stent grafts Follow-up outcomes. The median follow-up was
implanted at the initial TEVAR included those without 12.2 months (range, 3 days-8 years). Early causes of major
a proximal bare spring in 32 patients (Thoracic Excluder adverse events and death after the initial TEVAR are re-
stent grafts in 13 patients, redesigned and renamed “TAG” ported in Table IV. Prolonged reversible ischemic neuro-
stent grafts in 15, and TX2 stent grafts in four) and with logic deficit ($24 hours to #7 days) occurred in three
a proximal bare spring in nine (Valiant in eight and Talent patients. One patient required cerebrospinal fluid drainage.
in one). Deployment of two stent grafts was required in 11 The incidence of permanent paraplegia and renal insuffi-
patients at the initial procedure to increase the length of ciency requiring dialysis (>30 days) was 2% (n ¼ 1) and 0%,
coverage. Technical success, defined as sealing of the respectively. No late deaths (>30 days) were attributed to
primary entry tear without residual endoleak at the end of the primary TEVAR procedure.
the procedure, was achieved in all patients. One patient died of decompensated cirrhosis 20 days
Device diameters ranged from 28 to 46 mm. The mean after reintervention, and another patient died of a Stanford
proximal landing zone diameter was 30.73 mm (range, 23- type A aortic dissection 20 months after reintervention. No
36 mm), median proximal oversizing was 14.71% (range, renal or neurologic complication occurred.
8.8%-33.3%), and median length of aortic coverage was The overall mortality rate was 10% (n ¼ 4). The
15 cm (range, 9-31.5 cm). median survival time without reintervention was
Sixteen (39%) hybrid operations combining extra- 61.9 months (95% CI, 24.6-1) after TEVAR (Fig).
anatomic debranching of the supra-aortic vessels concomi- CT finding on follow-up. Complete false lumen
tant with TEVAR were required because of inadequate thrombosis at 1 year was 85% (n ¼ 35) in the region of
 JOURNAL OF VASCULAR SURGERY
330 Faure et al February 2014

Table IV. Early (<30 days) major outcomes after initial

endovascular thoracic aortic repair (TEVAR)

Secondary No
intervention reintervention Total
Outcome (n ¼ 13) (n ¼ 28) (n ¼ 41)

Death
Thoracic aortic 0 1 1
rupture
Mesenteric 0 1 1
ischemia
PRIND 2 1 3
Spinal cord injury 0 1 1
Renal failurea 2 1 3
Pulmonary failureb 1 1 2
Debranching 2 3 5
cervical access
site
complications

PRIND, Prolonged reversible ischemic neurologic deficit. Fig. Kaplan-Meier analysis for survival without reintervention for
a
Requiring dialysis. thoracic aortic endovascular repair (TEVAR) is shown with the
b
Requiring postoperative reintubation.
95% confidence interval (CI) (gray lines).

the stented aorta compared with 26% (n ¼ 10) distal to the Reinterventions consisted of 13 endovascular interven-
stented aorta. Complete false lumen thrombosis at the tions (93%): additional proximal endografting in nine
stent graft level was significantly higher in the “no reinter- (65%), requiring transposition of the supra-aortic vessel
vention” group (96% [n ¼ 25]) than in the “reinterven- in four (29%), and additional distal endografting in six
tion” group before reintervention (54% [n ¼ 7]; P ¼ (43%). Among them, two patients underwent additional
.003). At the end of the follow-up, complete false lumen proximal and distal endografting. One (2%) patient with
thrombosis at the stent graft level was 96% (n ¼ 25) in the a retrograde dissection underwent a composite graft
“no reintervention” group and 77% (n ¼ 10) after rein- replacement of the aortic valve, aortic root, and ascending
tervention in the “reintervention” group, but the differ- aorta, with reimplantation of the coronary arteries into the
ence was not significant. graft (Bentall procedure).
Including the whole study population, favorable aortic Technical success was achieved in 100%. Two deaths
remodeling, defined as stabilization or regression of the occurred after reintervention: one patient died of cirrhosis
false lumen diameter, was demonstrated in 39 patients at day 20 and another died of an ascending aortic dissection
(95%), and complete aortic remodeling, defined as at 20 months, with a reintervention-related mortality of 7%
complete reattachment of the dissecting membrane, was (n ¼ 1).
noted in four patients (10%). After the primary interven- Univariate logistic regression. Results of the univar-
tion, an increase in aortic diameter (>5 mm) in the region iate logistic regression analysis are reported in Table V. No
of the stented aorta occurred in eight patients (20%), statistically significant differences were found between the
which was attributable to type I endoleaks in two patients two groups in age, sex, hypertension, previous aortic
and to persistent retrograde flow in the false lumen surgery, length of the dissection (limited to the thoracic
through distal fenestrations in the other six. All of these aorta or extensive), proportion of acute or chronic dissec-
patients underwent a reintervention. After the secondary tion, length of the proximal landing zone, presence of
procedure, only one patient still had an increase in the transposition of the supra-aortic vessels, length of aortic
maximal aortic diameter despite a complete false lumen coverage, and false lumen thrombosis distal to the stent
thrombosis. graft. False lumen thrombosis at the stent graft level was
Reintervention. Fourteen reinterventions were per- significantly more frequent in patients not requiring rein-
formed in 13 patients (32%), the indications for which tervention than in those requiring reintervention before
included device migration in 2, proximal type I endoleak secondary procedure.
in 5, distal type I endoleak in 2, type II endoleak in 1, Multivariate logistic regression analysis. Multivar-
aneurysmal degeneration of the descending thoracic aorta iate analysis (Table V) demonstrated that proximal over-
below the stent graft in 1 and above the stent graft in 1, sizing $20% (OR, 16; 95% CI, 1.9-137.2; adjusted P ¼
and aneurysmal expansion of the dissected abdominal .011), the use of a stent graft with a proximal bare spring
aorta in 1. One patient required a third procedure for (OR, 12; 95% CI, 1.2-109.3; adjusted P ¼ .032), and
a retrograde dissection. The median delay to reinterven- long-term anticoagulant therapy (OR, 78%; CI, 1.5-999.3;
tion after implantation was 6.3 months (range, 7 days- adjusted P ¼ .03) were significant predictors for reinter-
7.3 years). vention after TEVAR for C-AD.
JOURNAL OF VASCULAR SURGERY
Volume 59, Number 2 Faure et al 331

Table V. Results of the univariate and multivariate logistic regressions

Univariate analysisa Multivariate analysisa,b

Variable OR (95% CI) P OR (95% CI) P

Age at first TEVAR (for each 5-year increase) 1.0 (0.8-1.3) .39 . .
Male sex 2.4 (0.4-18.2) .39 . .
Hypertension 0.6 (0.1-2.7) .51 . .
Anticoagulant therapy 37 (1.4-966.1) .03 78 (1.5-999.3) .03
Previous aortic surgery 2.4 (0.6-9) .21 . .
Extensive dissection 2 (0.2-15.5) .53 . .
Acute dissection 0.4 (0.1-1.9) .27 . .
Aneurysm diameter (for each 10-mm increase) 1.8 (1.1-3) .029 . .
Proximal landing zone length (for each 5-mm increase) 0.9 (0.6-1.3) .44 . .
Transposition of the supra-aortic vessels 1.0 (0.3-3.8) .98 . .
Stent graft with proximal bare spring 6.3 (1.3-31.3) .024 12 (1.2-109.3) .032
Length of aortic coverage (for each 20-mm increase) 1.0 (0.8-1.2) .95 . .
Proximal oversizing >20% 11 (2.3-55.9) .003 16 (1.9-137.2) .011
Postoperative false lumen thrombosis at stent graft level 70 (3.4-1) .006 . .
Postoperative false lumen thrombosis distal to the stent graft 3.4 (0.5-24.9) .23 . .

CI, Confidence interval; OR, odds ratio.

Risk factors for reintervention are indicated in bold.
a
Firth method logistic regression (penalizes likelihood).
b
Stepwise descending selection method.

DISCUSSION such as thoracic aortic disease or thoracic aortic degenera-

Since the first report of TEVAR repair in 1994 by Dake tive aneurysm.20 To the best of our knowledge, our study is
et al,10 indications for TEVAR have broadened. At present, the first to report the factors of reintervention after TEVAR
due to lower perioperative mortality and morbidity for aortic dissection.
compared with conventional open repair, the use of In our series of 41 elective patients, we report a 32%
TEVAR11,12 is not limited to treating degenerating aneu- reintervention rate at a mean follow-up of 4.7 years. These
rysms but is also used for complicated acute or chronic results are quite similar compared with those of the Med-
aortic dissection.13-15 TEVAR has reduced perioperative tronic Thoracic Endovascular Registry (MOTHER).6
mortality by more than two-thirds in the setting of aortic However, we report a higher secondary intervention rate
dissection but is associated with higher reintervention rates for chronic dissection (38%) than for acute dissection
than open repair and TEVAR for degenerative thoracic (20%), although this did not reach statistical significance
aneurysms. Long-term durability for this specific disease (P ¼ .2211). This is similar to the report by Böckler
remains unknown.2 et al18 but contradicts the finding of the MOTHER, where
Current published reports suggest that reintervention acute dissection had a higher reintervention rate of 54% vs
rates at up to 5 years seem to be higher for TEVAR for 29% at 6 years.6
C-AD (acute and chronic) than TEVAR for thoracic As reported in other studies, most secondary interven-
degenerative aneurysm. Parsa et al16,17 reported rates of tions did not require open surgical repair. In our patients,
23%, whereas Böckler et al18 reported rates of 32% in the all but one secondary intervention was managed by endo-
setting of C-AD, increasing to 46% at 5 years when vascular means. One patient was treated for a retrograde
TEVAR is performed for acute dissection.7 This compares ascendant aortic dissection. This is a lower rate than the
to reintervention rates of 15% to 17% after TEVAR for 80% open conversion rate for secondary procedure re-
thoracic aneurysm.5,19 Patterson et al6 recently reported ported by Neuhauser et al21 (all for type I endoleak) all
a lower mortality rate after TEVAR for patients with for retrograde dissection. Furthermore, no more complica-
chronic dissection (3%) than for patients with aneurysmal tions occurred after reintervention than after primary
disease (5%). The sample group reported higher reinterven- procedure, and the reintervention-related death rate was
tion rates for chronic (29%) or acute aortic dissection (54%) not higher. We do note that in our practice, the need for
at 6 years compared with TEVAR for thoracic aortic aneu- a hybrid procedure with open debranching is common in
rysm (16%).6 These results may be related to the higher the setting of secondary procedures (n ¼ 4 [31%]).
comorbidity rates in patients with thoracic aortic aneurysm Some authors have mentioned that most retrograde
than in patients treated for aortic dissection and confirm ascending aortic dissection is associated with the use of
that outcomes should be analyzed for specific pathologies proximal bare-spring stents for endograft treatment of
rather than for the type of procedure with the aim of dissection, with an incidence of 1.33% in the European
improving the durability of this technique. Registry on Endovascular Aortic Repair Complications
Several publications have reported factors associated and 2.48% for Dong et al22 and Eggebrecht H.23 Our
with reintervention after TEVAR for mixed indications previous study of retrograde ascending aortic dissection
 JOURNAL OF VASCULAR SURGERY
332 Faure et al February 2014

in the setting of surgical conversion after TEVAR high- cardiac reasons (auricular fibrillation or a mechanic valve)
lighted that aberrant subclavian arteries or aortic arch mal- should have closer follow-up.
formation and stent grafts with proximal bare springs seem A second significant anatomic factor for the long-term
to be associated with retrograde ascending aortic dissec- success of TEVAR appears to be the maximal diameter of
tion.24 In particular, careful analysis of our findings the dissected aorta, with a higher risk of reintervention in
revealed that the type of device used is a significant deter- case of dilatation leading to form aneurysms (P ¼ .002).
minant of TEVAR success in univariate and multivariate Along the same lines, some have suggested a lower
analysis (P ¼ .018). This contrasts with the Lee et al20 threshold than an aortic diameter of 5.5 cm for interven-
results for degenerating thoracic aneurysm, where the tion in type B dissection.29,30
type of device used was not significantly associated with As a consequence of the size of our study, the statistical
secondary intervention. analysis provided should be interpreted with caution. Our
This highlights differences in outcomes and complica- study points to the need for a specially designed prospective
tions according to the pathology being treated. On the registry on TEVAR complications for aortic dissection to
basis of our findings, the use of stent graft with proximal clarify the risk factors of reintervention in this specific area.
bare springs should be avoided in patients with aortic
dissection to decrease the rate of secondary procedures. CONCLUSIONS
Moreover, proximal oversizing $20% appears to be signif- This study confirms the short-term and midterm safety
icantly correlated with the need for reintervention and feasibility of TEVAR for complicated Stanford type B
(Table I). Most instructions for use of the current stent dissection. However, the reintervention rate is high
grafts recommend 10% to 20% oversizing with respect to (>30%). Excessive oversizing, the presence of a bare-
the preoperative aortic diameter; however, Dong et al25 spring stent in the proximal landing zone of the stent graft,
did not find significant differences in oversizing between dilation to form a large aneurysm, and anticoagulant
patients with and without stent graft-induced new entry therapy appear to be risk factors for reintervention.
after endovascular repair for Stanford type B aortic dissec- Complete false lumen thrombosis seems to be protective.
tion, and there is currently no consensus with regard to Anticipating these modifiable risk factors for reintervention
the optimal sizing of stent grafts for TEVAR in the setting appears essential for the long-term durability of TEVAR.
of aortic dissection.25
In our study of patients with reintervention, eight of AUTHOR CONTRIBUTIONS
the stent grafts placed had a proximal oversizing $20%,
Conception and design: EF
and the four type Ia endoleaks all developed in those
Analysis and interpretation: EF
patients with excessive oversizing. One other patient with
Data collection: EF, AC, RS
30% oversizing later required reintervention for device
Writing the article: EF
migration, and a further patient with 20% oversizing devel-
Critical revision of the article: LC, CM
oped a type Ia endoleak that spontaneously regressed. Our
Final approval of the article: PA
study suggests that oversizing should be <20% in the
Statistical analysis: LC
setting of TEVAR for type B dissection to reduce the risk
Obtained funding: GB
of complications such as dilatation of the aneurysm neck
Overall responsibility: PA
with subsequent migration of the endograft or type Ia
endoleak formation leading to reintervention.26
REFERENCES
Further analysis of the morphologic characteristics of
the dissections treated in our study showed that the stage 1. Dake MD, Kato N, Mitchell RS, Semba CP, Razavi MK, Shimono T,
et al. Endovascular stent-graft placement for the treatment of acute
(acute or chronic), extent of the dissection, length of aortic
aortic dissection. N Engl J Med 1999;340:1546-52.
coverage, length of the proximal landing zone even after 2. Tsai TT, Fattori R, Trimarchi S, Isselbacher E, Myrmel T,
debranching for adequate anchorage, and the status of Evangelista A, et al; International Registry of Acute Aortic Dissection.
the false lumen distal to the stent grafted aorta were not Long-term survival in patients presenting with type B acute aortic
significant factors leading to reintervention (Table V). dissection: insights from the International Registry of Acute Aortic
Dissection. Circulation 2006;114:2226-31.
However, complete false lumen thrombosis in the region 3. Suzuki T, Mehta RH, Ince H, Nagai R, Sakomura Y, Weber F, et al;
of the stent grafted aorta seems to be a significant anatomic International Registry of Aortic Dissection. Clinical profiles and
factor determining TEVAR long-term success (Table V). outcomes of acute type B aortic dissection in the current era: lessons
We have not shown that the status of the false lumen from the International Registry of Aortic Dissection (IRAD). Circula-
tion 2003;108:12-7.
thrombosis distally to the stent graft is a significant
4. Nienaber CA, Kische S, Ince H, Fattori R. Thoracic endovascular
predictor of outcomes, although most authors agree that aneurysm repair for complicated type B aortic dissection. J Vasc Surg
complete false lumen thrombosis is a protective factor pre- 2011;54:1529-33.
venting aortic growth and death.27,28 Owing to its effect 5. Leurs LJ, Harris PL, Buth J. Secondary interventions after elective
on thrombogenesis and therefore on false lumen throm- endovascular repair of degenerative thoracic aortic aneurysms: results of
the European collaborators registry (EUROSTAR). J Vasc Interv
bosis, long-term anticoagulant therapy is, unsurprisingly, Radiol 2007;18:491-5.
a significant factor in reintervention in this study. Patients 6. Patterson B, Holt P, Nienaber C, Cambria R, Fairman R,
who are receiving long-term anticoagulant therapy for Thompson M. Aortic pathology determines midterm outcome after
JOURNAL OF VASCULAR SURGERY
Volume 59, Number 2 Faure et al 333

endovascular repair of the thoracic aorta: report from the Medtronic of acute symptomatic and chronic expanding Stanford type B aortic
Thoracic Endovascular Registry (MOTHER) database. Circulation dissections. J Thorac Cardiovasc Surg 2006;132:361-8.
2013;127:24-32. 19. Desai ND, Burtch K, Moser W, Moeller P, Szeto WY, Pochettino A,
7. Duebener L, Hartmann F, Kurowski V, Richardt G, Geist V, Erasmi A, et al. Long-term comparison of thoracic endovascular aortic repair
et al. Surgical interventions after emergency endovascular stent-grafting (TEVAR) to open surgery for the treatment of thoracic aortic aneu-
for acute type B aortic dissections. Interact Cardiovasc Thorac Surg rysms. J Thorac Cardiovasc Surg 2012;144:604-9.
2007;6:288-92. 20. Lee CJ, Rodriguez HE, Kibbe MR, Chris Malaisrie S, Eskandari MK.
8. Arnaoutakis GJ, Hundt JA, Shah AS, Cameron DE, Black JH 3rd. Secondary interventions after elective thoracic endovascular aortic
Comparative analysis of hospital costs of open and endovascular repair for degenerative aneurysms. J Vasc Surg 2013;57:1269-74.
thoracic aortic repair. Vasc Endovascular Surg 2011;45:39-45. 21. Neuhauser B, Greiner A, Jaschke W, Chemelli A, Fraedrich G. Serious
9. Fillinger MF, Greenberg RK, McKinsey JF, Chaikof EL. Reporting complications following endovascular thoracic aortic stent-graft repair
standards for thoracic endovascular aortic repair (TEVAR). J Vasc Surg for type B dissection. Eur J Cardiothorac Surg 2008;33:58-63.
2010;52:1022-33. 1033.e15. 22. Eggebrecht H, Thompson M, Rousseau H, Czerny M, Lönn L,
10. Dake MD, Miller DC, Semba CP, Mitchell RS, Walker PJ, Liddell RP. Mehta RH, et al; European Registry on Endovascular Aortic Repair
Transluminal placement of endovascular stent-grafts for the treatment Complications. Retrograde ascending aortic dissection during or after
of descending thoracic aortic aneurysms. N Engl J Med 1994;331: thoracic aortic stent graft placement: insight from the European
1729-34. registry on endovascular aortic repair complications. Circulation
11. Brooks M, Loftus I, Morgan R, Thompson MJ. The Valiant thoracic 2009;120:S276-81.
endograft. Cardiovasc Surg (Torino) 2006;47:269-78. 23. Dong ZH, Fu WG, Wang YQ, Guo da Q, Xu X, Ji Y, et al. Retrograde
12. Makaroun MS, Dillavou ED, Wheatley GH, Cambria RP; Gore TAG type A aortic dissection after endovascular stent graft placement for
Investigators. Five-year results of endovascular treatment with the Gore treatment of type B dissection. Circulation 2009;119:735-41.
TAG device compared with open repair of thoracic aortic aneurysms. 24. Canaud L, Alric P, Gandet T, Albat B, Marty-Ané C, Berthet JP.
J Vasc Surg 2008;47:912-8. Surgical conversion after thoracic endovascular aortic repair. J Thorac
13. Lin PH, El Sayed HF, Kougias P, Zhou W, LeMaire SA, Coselli JS. Cardiovasc Surg 2011;142:1027-31.
Endovascular repair of thoracic aortic disease: overview of current 25. Dong Z, Fu W, Wang Y, Wang C, Yan Z, Guo D, et al. Stent graft-
devices and clinical results. Vascular 2007;15:179-90. induced new entry after endovascular repair for Stanford type B
14. Brown KE, Eskandari MK, Matsumura JS, Rodriguez H, aortic dissection. J Vasc Surg 2010;52:1450-7.
Morasch MD. Short and midterm results with minimally invasive 26. Jonker FH, Schlosser FJ, Geirsson A, Sumpio BE, Moll FL, Muhs BE.
endovascular repair of acute and chronic thoracic aortic pathology. Endograft collapse after thoracic endovascular aortic repair. J Endovasc
J Vasc Surg 2008;47:714-22. Ther 2010;17:725-34.
15. D’Souza S, Duncan A, Aguila F, Oderich G, Ricotta J, Kalra M, et al. 27. Trimarchi S, Tolenaar JL, Jonker FH, Murray B, Tsai TT, Eagle KA,
TEVAR for non-aneurysmal thoracic aortic pathology. Catheter Car- et al. Importance of false lumen thrombosis in type B aortic dissection
diovasc Interv 2009;74:783-6. prognosis. J Thorac Cardiovasc Surg 2013;145:S208-12.
16. Parsa CJ, Williams JB, Bhattacharya SD, Wolfe WG, Daneshmand MA, 28. Tsai TT, Evangelista A, Nienaber CA, Myrmel T, Meinhardt G,
McCann RL, et al. Midterm results with thoracic endovascular aortic Cooper JV, et al. Partial thrombosis of the false lumen in patients with
repair for chronic type B aortic dissection with associated aneurysm. acute type B aortic dissection. N Engl J Med 2007;357:349-59.
J Thorac Cardiovasc Surg 2011;141:322-7. 29. Lansman SL, Hagl C, Fink D, Galla JD, Spielvogel D, Ergin MA, et al.
17. Parsa CJ, Schroder JN, Daneshmand MA, McCann RL, Acute type B aortic dissection: surgical therapy. Ann Thorac Surg
Hughes GC. Midterm results for endovascular repair of complicated 2002;74:1833-5.
acute and chronic type B aortic dissection. Ann Thorac Surg 2010; 30. Crawford ES. The diagnosis and management of aortic dissection.
89:97-102. JAMA 1990;264:2537-41.
18. Böckler D, Schumacher H, Ganten M, von Tengg-Kobligk H,
Schwarzbach M, Fink C, et al. Complications after endovascular repair Submitted Jun 14, 2013; accepted Aug 28, 2013.