CLINICAL RESEARCH STUDIES

A randomized controlled trial of endovascular

aneurysm repair versus open surgery for
abdominal aortic aneurysms in low- to
moderate-risk patients
Jean-Pierre Becquemin, MD, Jean-Chistophe Pillet, MD, François Lescalie, MD, Marc Sapoval, MD,
Yann Goueffic, MD, Patrick Lermusiaux, MD, Eric Steinmetz, MD, and Jean Marzelle, MD, for the ACE
trialists, Creteil, France

Background: Several studies, including three randomized controlled trials (RCTs), have shown that endovascular repair
(EVAR) of abdominal aortic aneurysms (AAA) offered better early results than open surgical repair (OSR) but a similar
medium-term to long-term mortality and a higher incidence of reinterventions. Thus, the role of EVAR, most notably in
low-risk patients, remains debated.
Methods: The ACE (Anevrysme de l’aorte abdominale: Chirurgie versus Endoprothese) trial compared mortality and major
adverse events after EVAR and OSR in patients with AAA anatomically suitable for EVAR and at low-risk or intermediate-risk
for open surgery. A total of 316 patients with >5 cm aneurysms were randomized in institutions with proven expertise for both
treatments: 299 patients were available for analysis, and 149 were assigned to OSR and 150 to EVAR. Patients were monitored
for 5 years after treatment. Statistical analysis was by intention to treat.
Results: With a median follow-up of 3 years (range, 0-4.8 years), there was no difference in the cumulative survival free of death
or major events rates between OSR and EVAR: 95.9% ⴞ 1.6% vs 93.2% ⴞ 2.1% at 1 year and 85.1% ⴞ 4.5% vs 82.4% ⴞ 3.7%
at 3 years, respectively (P ⴝ .09). In-hospital mortality (0.6% vs 1.3%; P ⴝ 1.0), survival, and the percentage of minor
complications were not statistically different. In the EVAR group, however, the crude percentage of reintervention was higher
(2.4% vs 16%, P < .0001), with a trend toward a higher aneurysm-related mortality (0.7% vs 4%; P ⴝ .12).
Conclusions: In patients with low to intermediate risk factors, open repair of AAA is as safe as EVAR and remains a more
durable option. ( J Vasc Surg 2011;53:1167-73.)

Rupture of abdominal aortic aneurysm (AAA) is a fre- dii’s report raised hopes that endovascular repair of AAA
quent cause of cardiovascular death in industrialized coun- (EVAR) might improve outcomes.1 Meta-analysis of retro-
tries. Prophylactic open surgical repair (OSR) is indicated for spective studies as well as three prospective randomized con-
large asymptomatic AAA in patients with acceptable operative trolled studies (RCT) tended to confirm this hypothesis, at
risk. However, despite improvements in surgical expertise and least in the early stage.2-4
anesthesia, the death toll after OSR remains significant. Paro- After health care provider authorizations and stent graft
reimbursements, the number of patients undergoing OSR has
From the Department of Vascular Surgery, Hopital Henri Mondor, Univer-
sity Paris XII.
rapidly declined while the number undergoing EVAR has
This study was supported by a grant obtained from the French Ministry of expanded.5,6 However, rupture may still happen after EVAR,
Health that covered the cost of the study. The sponsor had no role in and reinterventions are not infrequent.3,4,7 As a consequence,
study design. the long-term efficacy of EVAR is still debated.
Competition of interest: J.P. Becquemin received fees for consulting and for
The ACE (Anevrysme de l’aorte abdominale: Chiru-
speaking from Cook, Medtronic, Gore, and Vascutek.
Additional materials for this article may be found online at www.jvascsurg. rgie versus Endoprothese) trial (http://ClinicalTrials.
org. gov, #NTC00224718) was conceived in 1998. This mul-
Reprint requests: Dr Jean-Pierre Becquemin, University of Paris, XII, VAS- ticenter, prospective randomized trial assessed the results of
CULAR SURGERY, Hopital Henri Mondor, University Paris XII 51 OSR vs EVAR in patients presenting with an asymptomatic
avenue duѧ 94000 Creteil, France (e-mail: jpbecquemin@hotmail.com).
The editors and reviewers of this article have no relevant financial relation-
AAA, deemed at low to moderate risk for surgery. We report
ships to disclose per the JVS policy that requires reviewers to decline the final results of this trial, with a median follow-up of 3 years.
review of any manuscript for which they may have a competition of
interest. METHODS
0741-5214/$36.00
Copyright © 2011 by the Society for Vascular Surgery. Participants. Inclusion criteria combined anatomic
doi:10.1016/j.jvs.2010.10.124 and clinical assessment:
1167
 JOURNAL OF VASCULAR SURGERY
1168 Becquemin et al May 2011

● Anatomic criteria were based on computed tomography formed at each evaluation. All events were collected in the
(CT) scan findings: AAA ⬎50 mm in men or ⬎45 mm in case report form.
women, or common iliac artery aneurysm ⬎30 mm and Study organization is summarized in Appendix 1 (online
upper neck free of major thrombus or calcifications and only). A multidisciplinary independent committee validated
at ⱖ15 mm in length and an angle between the neck and all end points.
the axis of the aneurysm ⬍60° and iliac arteries compat- Sample size. With the hypothesis of a 2-year 30%
ible with the introducer sheath. reduction of the incidence of death and major adverse
● Clinical assessment graded patients in categories 0 to 2 events in the EVAR group (␣, 5%; ␤ risk, 10%), 600
according to the comorbidity score of the Society for patients were required. Although the study was designed in
Vascular Surgery/American Association for Vascular 1999, it started in 2003 due to (1) conflicts between the
Surgery (SVS/AAVS).8 Ministry of Health and the National Healthcare Reim-
bursement Organization, (2) structural changes in regula-
Exclusion criteria were previous abdominal aortic sur- tory agencies and recommendations (temporary limitation
gery, ruptured aneurysm, mycotic aneurysm, severe iodine to high-risk patients), and (3) insurance issues between
allergy, life expectancy deemed ⬍6 months, or category 3 Clinical Research Direction and the centers. Because of a
of the SVS/AAVS classification. slow inclusion rate due to previously reported reasons,9 the
Centers fulfilled the recommendations issued in 2001 scientific committee stopped enrollment in March 2008
by the French Regulatory Agency for Medical Drug and and extended the follow-up up to 5 years. The trial ended in
Device Safety, which required a minimal activity of 20 AAA April 2009.
repairs/year and at least 8 EVAR procedures to be autho- Randomization. Patient data in each center were
rized to performed EVAR. By the time the study started, at checked by a multidisciplinary team (vascular surgeon, radiol-
least 30 EVAR procedures had been done in each center. ogist, and anesthesiologist). After written informed consent
Interventions. The protocol did not recommend any was obtained, the Clinical Research Unit of Henri Mondor
specific preoperative treatment (eg, ␤-blockers or statins). Hospital performed randomization stratified by center. Arm
allocation was notified ⱕ24 hours to the investigator.
Prophylactic antibiotics were injected intravenously at the
Statistical methods. Analysis was performed by
beginning of the procedure. Intravenous heparin (0.5 mg/
intention-to-treat. Patient characteristics, in-hospital post-
kg) was administered before clamping or sheath insertion.
operative data, and outcome measure are expressed as
Aspirin (75 mg/d) was given postoperatively to all patients.
means with standard deviation (SD) or as counts and
Any bifurcated or aortouniiliac stent graft with a Euro-
percentage. Qualitative variables were compared using ␹2
pean Conformity mark and approved by the French Health
test or Fisher test when appropriate. Quantitative variables
Security Agency could be used for EVAR. Access (cutdown
(mean ⫾ SD) were compared using the t test. Overall
or percutaneous) and management of hypogastric arteries
survival was calculated from the date of inclusion until
(embolization, overlap by stent graft) were left to the
death, major adverse complication, or last follow-up.
surgeon’s decision.
Event-free survival of minor complication was calculated
Surgical approach for transperitoneal or retroperitoneal from the date of inclusion until minor complication, death,
OSR, type of graft (polyester of polytetrafluoroethylene), major complication, or the last follow-up. Survival curves
combined revascularization of inferior mesenteric artery or hy- were estimated by the Kaplan-Meier method. Differences
pogastric artery, or both, were left to the surgeon’s decision. between the survival curves were tested for significance by
Outcomes. The primary end point was death of any the log-rank test. All P values reported were two-sided.
cause and major adverse events, including myocardial in- Differences between the results of comparative tests were
farction (defined by electrocardiogram signs, enzymes, and considered significant at P ⬍ .05. SPSS France 16.0 soft-
troponin elevation), permanent stroke, permanent hemo- ware (Bois Colombes, France) was used.
dialysis, major amputation, paraplegia, and bowel infarc-
tion. Secondary end points included vascular reinterven- RESULTS
tions and minor complications. Sexual impairment was Between March 2003 and March 2008, 25 centers
assessed at 1 year by asking for any alteration in sexual (Appendix 2) participated in the study. The mean number
function after treatment. No formal established question- of patients per center was 12 ⫾ 20 (range, 1-102).
naire was used. Buttock claudication was defined by the A flow diagram (Fig 1) shows results of the randomiza-
postoperative onset of a buttock pain that prevented a tion and protocol deviations. Of 306 randomized patients,
normal walking activity (ie, occurring for a walking distance 7 were excluded from analysis because of withdrawal of
of ⬍300 meters). Incisional complications included all wall consent. Intention-to-treat analysis was performed in 299
dehiscences and large abdominal wall palsy when the pa- patients (149 allocated to OSR and 150 to EVAR): 277
tient found it debilitating. Reinterventions for incisional (92.6%) were treated according to randomization, 1 patient
repair were not recorded. did not undergo intervention, and there were 21 cross-
Follow-up visits were scheduled at 1, 6, and 12 months, overs, mostly due to patient’s preferences. The crossover
and yearly thereafter. In the EVAR arm, plain abdominal rate was significantly higher in the OSR arm (11.4% vs
radiography, duplex scan, and contrast CT scan were per- 2.7%; P ⬍ .01). Mean time from randomization to treat-
JOURNAL OF VASCULAR SURGERY
Volume 53, Number 5 Becquemin et al 1169

Fig 1. Flow diagram of the study. EVAR, Endovascular aneurysm repair.

ment was 16.8 ⫾ 6.6 days (range, 1-29 days). Three Postoperative course. Table III summarizes the post-
patients were lost to follow-up in the EVAR group (2%), operative outcomes. Three deaths occurred postopera-
and five in the OSR group (3.3%). Patients were censored at tively, comprising two myocardial infarctions (one in each
the last available follow-up. group) and one cardiac arrest after immediate conversion to
Baseline data. There were 3 women (1.0%) and 296 OSR due to stent graft misdeployment. General complica-
men (99.0%). Mean age was 69 ⫾ 7 years (range, 49-83 tion and reintervention rates did not differ statistically. The
years). Preoperative risk factors, SVS/AAVS scores, American mean length of stay was 5.8 ⫾ 5.5 days after EVAR and
Society of Anesthesiologists classification, aneurysm classifica- 10.4 ⫾ 8.3 after OSR (P ⬍ .0001).
tion, and diameters are reported in Table I. Both groups were Long-term results. Median follow-up was 3 years
similar at baseline, with the exception of a significantly higher (mean 2.5 ⫾ 1.2 years; range, 0-4.8 years). Table IV
rate of category 2 coronary disease (16.8% vs 8.0%; P ⬍ .05) summarizes the details of deaths, major and minor adverse
and a more severe SVS/AAVS grading score (grade 2: 69.1% events, and reinterventions in the two groups.
vs 54.7%; P ⬍ .01) in the OSR arm. Survival. In addition to the 3 postoperative deaths,
Intervention. General anesthesia was used in all oper- 26 patients died during follow-up. There was no differ-
ations, except for eight patients in the EVAR arm (four ence in the cumulative survival rates between OSR and
local and four epidural anesthesia). EVAR according to the log-rank test (P ⫽ .24), with
In the OSR arm, surgery was performed through a 96.5% ⫾ 1.5% vs 95.2% ⫾ 1.8% at 1 year and 86.7% ⫾
midline incision in 70 patients, a transverse incision in 10, 4.4% vs 86.3% ⫾ 3.4% at 3 years, respectively.
and a retroperitoneal approach in 51. A tube graft was Deaths were related to the aneurysm or to the treatment
inserted in 70 patients and a bifurcated graft in 61. (aneurysm-related mortality) in one patient after OSR (0.6%)
In the EVAR arm, four types of stent grafts were used: and in six patients after EVAR (4%; P ⫽ .14). In the later case,
Zenith (Cook, Inc, Bloomington, Ind) in 81 patients (71 two deaths were reported as in the immediate postoperative
bifurcated, 10 aortouniiliac), Talent (Medtronic Vascular, course, two died of rupture, and the remaining two deaths
Santa Rose, Calif) in 52 (41 bifurcated, 11 aortouniiliac), occurred after vascular reinterventions.
Excluder (W. L. Gore and Assoc, Flagstaff, Ariz) in 9 (all Major adverse events. As shown in Fig 2, there was
bifurcated), and Powerlink (Endologix, Irvine, Calif) in 4 no difference in the cumulative survival free of death and
(all bifurcated). major adverse events rates between OSR and EVAR, with
Operative details are summarized in Table II. EVAR 95.9% ⫾ 1.6% vs 93.2% ⫾ 2.1% at 1 year, and 85.1% ⫾ 4.5%
was associated with a shorter duration of intervention and vs 82.4% ⫾ 3.7% at 3 years, respectively (P ⫽ .09). The rates
of ventilatory support, and less blood transfusion. As ex- of myocardial infarction, neurologic complications, and
pected, X-ray exposure time and amount of contrast were renal failure were in the same range in the two groups.
higher with EVAR. There were three ruptures in the EVAR group, all ⬎2 years
 JOURNAL OF VASCULAR SURGERY
1170 Becquemin et al May 2011

Table I. Patient characteristics Table III. In-hospital postoperative data

Open repair EVAR Open repair EVAR

Variablea (n ⫽ 149) (n ⫽ 150) Variablea (n ⫽ 149) (n ⫽ 150) P

Risk factors Length of stay, days 10.4 ⫾ 8.3 5.8 ⫾ 5.5 ⬍.0001
Age, years 70 ⫾ 7.1 68.9 ⫾ 7.7 Median 8.0 5.0 ...
Range 54-83 49-83 30-day deaths (all causes) 1 (0.6) 2 (1.3) NS
Male sex 146 (98) 151 (100) Major adverse events,
Diabetes 29 (19.5) 20 (13.3) patientsb 1 (0.6) 3 (2) NS
Tobacco 74 (49.7) 73 (48.7) Acute MI 1 1 ...
Hypertension 95 (63.8) 99 (66.0) Paraplegia ... 1 ...
Hyperlipidemia 98 (65.8) 103 (68.7) Renal failure ... 1 ...
Carotid artery disease 12 (8.1) 12 (8.0) Stroke ... 1b ...
Coronary disease Reinterventions, patients 2 (1.3) 8 (5.3) NS
0 84 (56.4) 101 (67.3)
1 40 (26.8) 37 (24.3) EVAR, Endovascular aneurysm repair; MI, myocardial infarction.
2 25 (16.8) 12 (8.0)
a
Continuous data are shown as mean ⫾ standard deviation and median;
Renal insufficiency 15 (10.1) 21 (14) categoric data are number (%).
b
Pulmonary disease 42 (28.2) 29 (19.3) Five complications occurred in 4 patients in the EVAR arm, and this patient
SVS/AAVS risk 1.65 ⫾ 0.51 1.51 ⫾ 0.53 also had renal failure.
ASA classification
1 12 16
2 89 99 Table IV. All outcome measures in study patients
3 48 34
4 — 2 Open repair EVAR
AAA classification (n ⫽ 149) (n ⫽ 150)
Type A 48 50 Variable No. (%) No. (%) P
Type B 52 56
Type C 16 17 Deaths 12 (8) 17 (11.3) NS
Type D 21 17 Major adverse events 6 (4) 10 (6.7) NS
Type E 4 6 Stroke 1 (0.7) 1 (0.7) NS
Type F ⫽ saccular 1 — Myocardial infarction 4 (2.7) 6 (4) NS
Type G ⫽ iliac 6 5 Paraplegia ... 1 (0.7) NS
AAA diameter,b mm 55.6 ⫾ 6.6 55.2 ⫾ 8.1 Renal failure 1 (0.7) 3 (2) NS
AAA rupture ... 3 (2.0) NS
AAA, Abdominal aortic aneurysm; ASA, American Society of Anesthesiol- Reinterventions 4 (2.7) 24 (16) ⬍.0001
ogists; EVAR, endovascular aneurysm repair; SVS/AAVS, Society for Vas- Minor adverse events 73 (48.7) 62 (41.3) NS
cular Surgery/American Association for Vascular Surgery. Hemorrhage 7 (4.7) 4 (2.7) NS
a
Categoric data are presented as number (%) and continuous data as Infection 11 (7.4) 14 (9.3) NS
means ⫾ standard deviation and as noted. Minor cardiac complications 19 (12.8) 9 (6) ⬍.05
b
Excluding 11 iliac aneurysms. Respiratory complications 8 (5.4) 5 (3.3) NS
Atheroembolism 1 (0.7) 3 (2) NS
Graft infection 1 (0.7) ... NS
Table II. Details of aneurysm repair Lymphorrhea/lymphocele ... 4 (2.7) NS
Incisional complications 38 (25.5) 1 (0.7) ⬍.0001
Open repair EVAR Buttock claudication 3 (2) 21 (14) ⬍.001
Variablea (n ⫽ 149) (n ⫽ 150) P Sexual dysfunction 11 (7.4) 7 (4.7) NS

Length of
intervention,
hours 2.8 ⫾ 1.1 2.1 ⫾ 0.9 ⬍.0001 graft and the recipient iliac artery. All three patients under-
X-ray exposure time,
went emergency aortic repair: one patient died postopera-
minutes 1.9 ⫾ 7.2 16.3 ⫾ 13.5 ⬍.0001
Contrast volume, mL 13 ⫾ 46 131 ⫾ 101 ⬍.0001 tively, one patient died at 4 months, and one patient
Ventilatory support, 8 ⫾ 13.7 3.2 ⫾ 2.5 ⬍.0001 survived.
hours Reinterventions. Fig 3 shows that there was a signif-
Median 5.5 3.0 icant difference in the cumulative survival free of death and
RBC transfusion, U 2.1 ⫾ 4.0 0.2 ⫾ 0.9 ⬍.0001
Median 2.0 0 vascular reintervention rates—including graft replacement
and endovascular or open repair of endoleaks, occlusions or
EVAR, Endovascular aneurysm repair; RBC, red blood cells. stenoses— between OSR and EVAR, with 96.5% ⫾ 1.5% vs
a
Data are presented as mean ⫾ standard deviation, unless otherwise indi-
cated.
91.3% ⫾ 2.3% at 1 year and 85.8% ⫾ 4.5% vs 76.1% ⫾ 4.6%
at 3 years, respectively (P ⫽ .01). In the EVAR group, the
crude percentage of vascular reintervention rate was higher
after EVAR: two patients had documented endoleaks (one (2.7% vs 16%, P ⬍ .0001) with a trend toward a higher
type I and one type II) which were awaiting treatment, and aneurysm-related mortality (0.7% vs 4%; P ⫽ .12).
one patient ruptured 2 months after a normal CT scan due Minor complications. There was no significant differ-
to a sudden disconnection between the distal limb of the ence in minor complications between OSR and EVAR.
JOURNAL OF VASCULAR SURGERY
Volume 53, Number 5 Becquemin et al 1171

aneurysms, in whom the hypogastric artery was coil embo-

lized or lost (14 of 45 [31%]).
Endoleaks. CT scan found endoleaks in 41 of 150
patients (27%) in the EVAR arm. Among 10 type I en-
doleaks, 2 were treated by open surgery and graft replace-
ment, 5 were successfully treated by an endoluminal pro-
cedure, and 3 were awaiting treatment at the time of this
analysis. Of 31 type II endoleaks, 8 were treated by coil
embolization, and 23 were left untreated.
Sexual function. Sexual assessment 1 year after treat-
ment showed no difference between the two treatments
(Table IV), although there was a trend toward more sexual
dysfunction in the OSR group.
Given the crossover rate, a per-protocol analysis was
performed but did not change the conclusions.

DISCUSSION
The ACE trial shows that EVAR or OSR in low-risk to
moderate-risk patients carries a similar risk of early-term
and medium-term (up to 4.8 years) death, major adverse
events, and minor complications. However, EVAR was
Fig 2. Kaplan-Meier curve of survival free of death or major associated with more vascular reinterventions and a trend
events after open surgical repair (OSR) or endovascular aneurysm
toward higher aneurysm-related mortality.
repair (EVAR).
Three RCTs2-4 have reported lower 1-month postop-
erative mortality rates after EVAR compared with OSR
(difference range, 2.5%-4.1%). Accordingly, the propensity
score-matched analysis of a 45,660 cohort of Medicare
patients confirmed a reduction in the postoperative mortal-
ity after EVAR (1.2% vs 4.8%).10 In ACE, we did not find
such a reduction after EVAR: the postoperative mortality
rate was very low in the OSR group, whereas mortality after
EVAR was in the same range as the three previous RCTs
(0.5% to 2.1%). After OSR, a similar low mortality rate of
3% was found in Veterans Affairs Open versus Endovascular
Repair (OVER) trial,2 in control groups of the U.S. Food
and Drug Administration phase 2 trials (range, 0%-
2.7%),11-13 and in centers of excellence,14,15 but the rate
was 4.1% in Dutch Randomised Endovascular Aneurysm
Management (DREAM)4 trial and 6.2% in the Comparison
of Endovascular Aneurysm Repair with Open Repair in
Patients with Abdominal Aortic Aneurysm (EVAR 1) trial.3
Among numerous factors potentially accounting for dif-
ferences in mortality rates, study design, population studied,
selection of centers, national standards of care, and date of
publication may be relevant. Although most patients in the
three previous RCTs were fit for surgery, baseline characteris-
Fig 3. Kaplan-Meier curve of survival free of death or reinterven-
tics may not be fully identical, and a similar distribution of risk
tion after open surgical repair (OSR) or endovascular aneurysm
repair (EVAR). factors does not fully take into account the association of risks.
The OVER trial used the RAND surgical risk score, with only
53% patients categorized as at low risk for surgery. Patients in
EVAR 1 or DREAM may have been at higher risk for OSR
However, the OSR group experienced more minor cardiac because risk assessment was left to each center team’s appre-
and incisional complications, whereas buttock claudication ciation.
was more frequent in the EVAR group. Of the incisional We chose the SVS/AAVS grading system. Although it
complications in the OSR group, 36.3% (29 of 80) oc- may be less accurate than scoring systems such as the Physio-
curred after a transperitoneal approach and 17.6% (9 of 51) logical and Operative Severity Score for enUmeration of Mor-
after a retroperitoneal approach. Buttock claudication was tality and Morbidity (POSSUM) score,16 Glasgow Aneurysm
mostly observed in patients with types C, D, E, or G Score,17 or the fitness score proposed by the EVAR-1 trial-
 JOURNAL OF VASCULAR SURGERY
1172 Becquemin et al May 2011

ists,18 it may have excluded severe risk factors such as renal, who died after EVAR was considered but not proven.
cardiac, or pulmonary insufficiencies more efficiently than EVAR-1 reported 13 ruptures: 10 before treatment, and 3
previous trials. As in the OVER trial, and by hazard of enroll- after treatment (2 after EVAR, 1 after OSR).
ment, fewer women were enrolled which may affect the gen- The distribution of recorded complications differed ac-
eralization of the results. More favorable results after OSR cording to treatment: more buttock claudication occurred
may also be explained by the size of aneurysms in the ACE after EVAR and more incisional complications after OSR.
trial, smaller than EVAR 1, but similar to OVER and Buttock claudication is a well-known complication of over-
DREAM. However, results of the OVER trial suggest a very lapping hypogastric arteries by the limb of a stent graft
limited role of AAA diameter on the early outcome in the when the aneurysm extends close to or involves the
population studied in those RCTs. Finally, the favorable anat- hypogastric bifurcation. Although spontaneous im-
omy for EVAR of all randomized patients also means an easier provement can be expected over time, we previously
and more straightforward open surgical repair. reported that 15% of patients remain disabled despite
Volume and qualification of centers are important to rehabilitation.26 The patency of the hypogastric artery
successfully deal with postoperative complications.19 The vol- should be maintained, whenever feasible, in healthy pa-
ume threshold is about 30 cases/year.20,21 A review of 3912 tients, especially those who have normal walking activity,
patients undergoing AAA repair22 found postoperative mor- such as most of the patients in this trial. At present, this
tality was 2.2% for vascular surgeons, 4% for cardiac surgeons, may be achieved by inserting currently available large
and 5.5% for general surgeons (P ⬍ .001). In the frame of distal limbs in the common iliac artery or by using
RCTs, it is difficult to compare centers: in ACE, as in OVER, branched stent grafts in selected cases.27
OSR was performed by full-time vascular surgeons, but in the Incisional complications are frequent after OSR and seem
EVAR 1 trial, although a few highly trained centers performed more frequent in patients with AAA than in patients with
EVAR, OSR was also performed by general surgeons. occlusive disease.28 In the OVER trial, 5% of patients in the
DREAM and EVAR-1 trialists enrolled patients between OSR group underwent incision hernia repair. Similarly to
1999 and 2003, OVER started in 2002, and ACE in 2003: we EVAR-1 and DREAM, we did not record specifically all
cannot exclude that improvement in patients’ preoperative, abdominal wall repairs or the incidence of bowel obstruction.
intraoperative, and postoperative management affected mor- That 24% of the patients in our OSR group presented inci-
tality reduction. sional complications underlines one drawback of the open
As reported in previous RCTs and defined in practice approach. Finally, the higher rate of minor cardiac complica-
guidelines,23 blood loss, need for transfusion, duration of tions in the OSR group may be related to a higher rate of
postoperative mechanical ventilation, and hospital length of coronary disease in this group or to the more invasive proce-
stay were significantly lower in the EVAR group. However, dure, or both.
we did not find a significant reduction of postoperative general Similar to the OVER trial, where 12% of EVAR patients
complication rates between the two treatments. Of note, the underwent a vascular reintervention vs 1.6% in the OSR
length of stay after both techniques was longer than usually group, in the ACE trial, vascular reinterventions occurred in
reported in the United States. This had more to do with the 16% of the EVAR group vs 2.7% in the OSR group. In
difference in the health care system and behaviors of patients EVAR-1, the overall rates of graft-related complications and
and physicians than with the occurrence of complications. The reinterventions were higher by a factor of three to four in the
length of stay was much shorter with EVAR. endovascular group. However, indications for reintervention
Survival, with a median follow-up of 3 years (up to 4.8 were highly variable: rupture, thrombosis, or type I endoleak
years) did not differ between the two treatments options. were indications for reintervention, whereas the need for
These results, consistent with the 2-year results of the reintervention in type II endoleaks was still debated. The
DREAM4 and OVER2 trials and the 4-year results of DREAM24 investigators rightly stated that reintervention
EVAR-1,3 are confirmed by recent reports of, respectively, 6 is a “soft” end point because the decision is at the surgeon’s
and 8 years of follow-up for DREAM24 and EVAR-1.25 The discretion.
expected less invasiveness of EVAR did not translate into a Finally the lack of significant difference in sexual function
significant reduction of death or of major adverse events, impairment between OSR and EVAR in ACE trial is consis-
which is also in agreement with the OVER trial findings. In tent with the OVER2 and DREAM29 findings.
other words, low-risk and intermediate-risk patients sustain One weakness of this study was the slow pace of enroll-
well the more aggressive treatment of OSR. ment and the failure to reach the expected number of patients.
AAA ruptures remain the Achilles’ heel of EVAR. Despite This may have affected the power of the analysis, as may have
a thorough follow-up, three ruptures (2%) occurred in the been the case in the DREAM trial, which was of a similar
EVAR arm, leading to two deaths despite emergency inter- magnitude. Except for the early mortality rates, however, our
vention. Two ruptures may have been prevented by more findings were in agreement with the three previous RCTs. A
expedient treatment of identified endoleaks and in the third meta-analysis of these four trials may help to clarify the relative
patient by a longer limb overlap in the common iliac artery. indication of EVAR and open surgery in AAA patients.
The OVER trial did not report any rupture, but follow-up was Finally, these results, as those of previous trials, reflect
shorter (1.8 years). The DREAM24 trial reported one rupture current stent graft technology. With continuous advances in
before treatment, and the possibility of rupture in two patients stent graft design, it may be possible, as shown by the Euro-
JOURNAL OF VASCULAR SURGERY
Volume 53, Number 5 Becquemin et al 1173

pean Collaborators on Stent-Graft Techniques for AAA and 11. Greenberg RK, Chuter TA, Sternbergh WC, Fearnot NE; Zenith
Thoracic Aortic Aneurysm and Dissection Repair (EURO- Investigators. Zenith AAA endovascular graft: intermediate-term re-
sults of the US multicenter trial. J Vasc Surg 2004;39:1209-18.
STAR) results,30 that the durability of EVAR will improve in 12. Moore WS, Matsumura JS, Makaroun MS, Katzen BT, Deaton DH, Decker
the future. M, et al. Five-year interim comparison of the Guidant bifurcated endograft
with open repair of abdominal aortic aneurysm. J Vasc Surg 2003;38:46-55.
CONCLUSIONS 13. Zarins CK, White RA, Schwarten D, Kinney E, Diethrich EB, Hodgson
KJ, et al. AneuRx stent graft versus open surgical repair of abdominal
In a selected group of patients with low to intermediate aortic aneurysms: multicenter prospective clinical trial. J Vasc Surg
risk factors, OSR and EVAR offer no difference in survival or 1999;29:292-305.
in major and minor complications. The choice between OSR 14. Conrad MF, Crawford RS, Pedraza JD, Brewster DC, LaMuraglia GM,
Corey M, et al. Long-term durability of open abdominal aortic aneu-
and EVAR should rely on the balance of different risks: more
rysm repair. J Vasc Surg 2007;46:669-75.
postoperative transfusions, a longer hospital stay, and inci- 15. Hertzer NR, Mascha EJ, Karafa MT, O’Hara PJ, Krajewski LP, Beven
sional complications with OSR vs the need of follow-up with EG. Open infrarenal abdominal aortic aneurysm repair: the Cleveland
repeat CT scans, a higher rate of vascular reinterventions, and Clinic experience from 1989 to 1998. J Vasc Surg 2002;35:1145-54.
a small but persistent risk of rupture with EVAR. 16. Neary WD, Crow P, Foy C, Prytherch D, Heather BP, Earnshaw JJ.
Comparison of POSSUM scoring and the Hardman Index in selection
of patients for repair of ruptured abdominal aortic aneurysm. Br J Surg
AUTHOR CONTRIBUTIONS 2003;90:421-5.
Conception and design: JPB, MS 17. Baas AF, Janssen KJ, Prinssen M, Buskens E, Blankensteijn JD. The
Glasgow Aneurysm Score as a tool to predict 30-day and 2-year mor-
Analysis and interpretation: JPB JM tality in the patients from the Dutch Randomized Endovascular Aneu-
Data collection: JPB JM, JCP, FL, YG, PL, ES rysm management trial. J Vasc Surg 2008;47:277-81.
Writing the article: JPB JM 18. Brown LC, Greenhalgh RM, Howell S, Powell JT, Thompson SG.
Critical revision of the article: JPB JM Patient fitness and survival after abdominal aortic aneurysm repair in
patients from the UK EVAR trials. Br J Surg 2007;94:709-16.
Final approval of the article: JPB, JM, MS, JCP, FL, YG,
19. Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality
PL, ES associated with inpatient surgery. N Engl J Med 2009;361:1368-75.
Statistical analysis: JPB, JM 20. Birkmeyer JD, Siewers AE, Finlayson EV, Stukel TA, Lucas FL, Batista
Obtained funding: JPB MS I, et al. Hospital volume and surgical mortality in the United States.
Overall responsibility: JPB N Engl J Med 2002;346:1128-37.
21. Holt PJ, Poloniecki JD, Khalid U, Hinchliffe RJ, Loftus IM, Thompson MM.
Effect of endovascular aneurysm repair on the volume-outcome relationship in
aneurysm repair. Circ Cardiovasc Qual Outcomes 2009;2:624-32.
REFERENCES
22. Dimick JB, Cowan JA Jr, Stanley JC, Henke PK, Pronovost PJ, Up-
1. Parodi JC, Palmaz JC, Barone HD. Transfemoral intraluminal graft implanta- church GR. Surgeon speciality and provider volumes are related to
tion for abdominal aortic aneurysms. Ann Vasc Surg 1991;5:491-9. outcome of intact abdominal aortic aneurysm repair in the united
2. Lederle FA, Freischlag JA, Kyriakides TC, Padberg FT, Matsumura JS, States. J Vasc Surg 2003;38:739-44.
Kohler TR, et al. Outcomes following endovascular vs open repair of 23. Chaikof EL, Brewster DC, Dalman RL, Makaroun MS, Illig KA, Sicard
abdominal aortic aneurysm: a randomized trial. JAMA 2009;302:1535-42. GA, et al. The care of patients with an abdominal aortic aneurysm: the
3. EVAR trial participants. Endovascular aneurysm repair versus open Society for Vascular Surgery practice guidelines. J Vasc Surg 2009;50(4
repair in patients with abdominal aortic aneurysm (EVAR trial 1): Suppl):S2-49.
randomised controlled trial. Lancet 2005;365:2179-86. 24. De Bruin JL, Baas AF, Buth J, Prinssen J, Verhoeven ELG, Cuypers
4. Blankensteijn JD, de Jong SE, Prinssen M, van der Ham AC, Buth J, van PWM, et al. Long-term outcome of open or endovascular repair of
Sterkenburg SMM, et al; Dutch Randomized Endovascular Aneurysm abdominal aortic aneurysm. N Engl J Med 2010;362:1881-9.
Management (DREAM) Trial Group. Two-year outcomes after con- 25. The United Kingdom EVAR Trial Investigators. Endovascular versus Open
ventional or endovascular repair of abdominal aortic aneurysms. N Engl Repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1863-71.
J Med 2005;352:2398-405. 26. Farahmand P, Becquemin JP, Desgranges P, Allaire E, Marzelle J,
5. Giles KA, Pomposelli F, Hamdan A, Wyers M, Jhaveri A, Schermerhorn Roudot-Thoraval F. Is hypogastric artery embolization during endovas-
ML. Decrease in total aneurysm-related deaths in the era of endovascu- cular aortoiliac aneurysm repair (EVAR) innocuous and useful? Eur J
lar aneurysm repair. J Vasc Surg 2009;49:543-50. Vasc Endovasc Surg 2008;35:429-35.
6. Anderson PL, Arons RR, Moskowitz AJ, Gelijns A, Magnell C, Faries 27. Minion DJ, Xenos E, Sorial E, Saha S, Endean ED. The trifurcated
PL, et al. A statewide experience with endovascular abdominal aortic endograft technique for hypogastric preservation during endovascular
aneurysm repair: rapid diffusion with excellent early results. J Vasc Surg aneurysm repair. J Vasc Surg 2008;47:658-61.
2004;39:10-9. 28. Raffetto JD, Cheung Y, Fisher JB, Cantelmo NL, Watkins MT, Lamorte WW,
7. Harris PL, Vallabhaneni SR, Desgranges P, Becquemin JP, van Marrewijk et al. Incision and abdominal wall hernias in patients with aneurysm or occlu-
C, Laheij RJ, et al. Incidence and risk factors of late rupture, conversion, sive aortic disease. J Vasc Surg 2003;37:1150-4.
and death after endovascular repair of infrarenal aortic aneurysms: the 29. Prinssen M, Buskens E, Nolthenius RP, van Sterkenburg SM, Teijink
EUROSTAR experience. European Collaborators on stent/graft tech- JA, Blankensteijn JD. Sexual dysfunction after conventional and endo-
niques for aortic aneurysm repair. J Vasc Surg 2000;32:739-49. vascular AAA repair: results of the DREAM trial. J Endovasc Ther
8. Chaikof EL, Fillinger MF, Matsumura JS, Rutherford RB, White GH, Blan- 2004;11:613-20.
kensteijn JD, et al. Identifying and grading factors that modify the outcome of 30. Hobo R, Buth J. Secondary interventions following endovascular ab-
endovascular aortic aneurysm repair. J Vasc Surg 2002;35:1061-6. dominal aortic aneurysm repair using current endografts. A Eurostar
9. Becquemin JP. The ACE trial: a randomized comparison of open versus report. J Vasc Surg 2006;43:896-902.
endovascular repair in good risk patients with abdominal aortic aneu-
rysm. J Vasc Surg 2009;50:222-4. Submitted Sep 1, 2010; accepted Oct 26, 2010.
10. Schermerhorn ML, O’Malley AJ, Jhaveri A, Cotterill P, Pomposelli F,
Landon B. Endovascular vs. open repair of abdominal aortic aneurysms Additional materials for this article may be found online
in the Medicare population. N Engl J Med 2008;358:464-74. at www.jvascsurg.org.
 JOURNAL OF VASCULAR SURGERY
1173.e1 Becquemin et al May 2011

APPENDIX 1 (ONLINE ONLY). Hôpital Trousseau, Tours: Patrick Lermusiaux, Robert

ACE study committees and organization Martinez, François Bleuet, Dupreix (n ⫽ 17);
CHU de Pontchaillou, Rennes: Jean Philippe Verhoye,
Writing committee: Jean-Pierre Becquemin (correspond- Thierry Langanay, Jean François Heautot (n ⫽ 2);
ing author), Jean Marzelle
Hôpital Cardiologique, Lille: Mohamad Koussa, Stephan
Principal investigators: Jean-Pierre Becquemin, Marc Sa-
Haulon, Pascal Halna, Laurence Destrieux, Christophe
poval
Lions, Serge Wiloteaux, Jean Paul Beregi (n ⫽ 5);
Scientific committee: Jean-Pierre Becquemin, Jean-Pierre
Hôpital Saint Joseph, Marseille: Patrice Bergeron, Jean-
Favre, Jacques Watelet, Patrick Lermusiaux, Marc Sapoval
Jacques Pinot (n ⫽ 11);
Statistics: Eric Lepage, François Hemery, Guillaume Dol-
Hôpital G et R Laennec, Nantes: Philippe Patra, Alain Co-
beau, Nasser Hawajry, Patric Cunin
Surveillance committee: Peter Harris, Luc Stockx, Gilles stargent, Philippe Chaillou, Aline D’Alicourt, Yann Goueffic
Chatellier (n ⫽ 20);
Safety issues committee: Claude Mialhe, Jean-Noel Centre Hospitalier René Dubos, Pontoise: Eric Cheysson,
Fiessinger, Luc Pagny, Hicham Kobeiter Alain Parrot, Patrick Garance, Icham Abada (n ⫽ 1);
End points and adverse events validation committee: CH de Valenciennes, Valenciennes: Arnaud Demon, Ab-
Christian Boissier, Philippe Lacroix, François Ledru, delhakim Tyazi (n ⫽ 4);
Jean-Jacques Pinot Jean-François Deux, Boyan Tzvet- Nouvelles Cliniques Nantaises, Nantes: Jean-Christophe
kov, Philippe Duvaldestin, Jacques Watelet Pillet (n ⫽ 33);
Data management: Cécile Jourdain, Virginie David, Del- Clinique Saint-Augustin, Nantes: François Lescalie, Gé-
phine Enouf, Namou Ady, Amor Krimi, Noël Boudjema rard Tilly (n ⫽ 25);
Hôpital du Bocage, Dijon: Eric Steinmetz, Claire Favier,
APPENDIX 2 (ONLINE ONLY). Roger Brenot, Denis Krause, Jean Pierre Cercueil (n ⫽ 15);
ACE trial participating centers and physicians with Clinique Pasteur, Toulouse: Olivier Vahdat, Michel Sauer,
corresponding number of cases. Philippe Soula, Aristide Querian, Olivier Garcia, Michel
Levade, Daniel Colombier (n ⫽ 2);
CHU AngersⴚHôtel Dieu, Angers: Yann Jousset, Bernard
Clinique des Franciscaines, Nimes: Jean-Marie Cardon, An-
Enon, Vincent Blin, Jean Picquet, Philippe L’Hoste, Fran-
dré Joyeux, Pierre Borrelly, Georges Dogas (n ⫽ 10);
cine Thouveny (n ⫽ 3);
Hôpital d’adultes de la Timone, Marseille: Pierre-Édouard
Hôpital privé d’Antony, Antony: Hervé Borie, Stéphane
Kowarski, Jean-Marc Pernes, Mario Auguste (n ⫽ 5); Magnan, Alain Branchereau, Jean-Michel Bartoli (n ⫽ 1);
Hôpital Henri Mondor, Creteil: Jean-Pierre Becquemin, Hôpital Saint Roch: Nice, Réda Hassen-Khodja, Michel
Pascal Desgranges, Eric Allaire, Jean Marzelle, Hicham Batt, Pierre-Franck Planchard, Pierre-Jean Bouillanne,
Kobeiter (n ⫽ 102); Pierre Haudebourg, Jean Bayne (n ⫽ 10);
Clinique mutualiste des Eaux Claires, Grenoble: Pierre- Hôpital La Cavale Blanche, Brest: Pierre Gouny, Ali
Yves Meaulle, Dominique Chaix (n ⫽ 1); Badra, Jacques Braesco, Michel Nonent (n ⫽ 2);
Hôpital Européen Georges Pompidou, Paris: Pierre Juliae, Hôpital Sud, Rennes: Antoine Lucas, Alain Cardon, Yvon
Jean Noël Fabiani, Patrick Chevalier, Myriam Combes, Ag- Kerdiles, Yann Rolland (n ⫽ 3);
athe Seguin, Denis Belhomme, Marc Sapoval, Jean Baque, Polyclinique de Poitiers, Poitiers: Michael Kassab,
Olivier Pellerin (n ⫽ 3); Christophe Brillu, François Goubault, Laurent Tailboux
CHU Nord, Saint Étienne: Jean Pierre Favre, Xavier Bar- (n ⫽ 2);
ral, Charles Veyret (n ⫽ 14); Clinique Marie Immaculée, Bourges:, Hugues Darrieux,
Hôpital Charles Nicolle, Rouen: Jacques Watelet, Chris- Olivier Briand, Jean-Claude Maillard (n ⫽ 6).
tophe Peillon, Didier Plissonier, Pascale Thomas, Eric Cla-
vier (n ⫽ 9);