- Email: [email protected]
Reintervention Rate after Open Surgery and Endovascular Repair for Nonruptured Abdominal Aortic Aneurysms
Accepted Manuscript Reintervention Rate after Open Surgery and Endovascular Repair for Nonruptured Abdominal Aortic Aneurysms Deokbi Hwang, Sujin Park, Hyung-Kee Kim, Jong-Min Lee, Seung Huh PII:
S0890-5096(16)30698-7
DOI:
10.1016/j.avsg.2017.03.168
Reference:
AVSG 3278
To appear in:
Annals of Vascular Surgery
Received Date: 21 August 2016 Revised Date:
2 February 2017
Accepted Date: 15 March 2017
Please cite this article as: Hwang D, Park S, Kim HK, Lee JM, Huh S, Reintervention Rate after Open Surgery and Endovascular Repair for Nonruptured Abdominal Aortic Aneurysms, Annals of Vascular Surgery (2017), doi: 10.1016/j.avsg.2017.03.168. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Reintervention Rate after Open Surgery and Endovascular Repair for
2
Nonruptured Abdominal Aortic Aneurysms
4
RI PT
3
Deokbi Hwanga, Sujin Parka, Hyung-Kee Kima, Jong-Min Leeb, and Seung Huha
7
a
Division of Vascular Surgery, Department of Surgery, bDepartment of Radiology, Kyungpook National University School of Medicine, Daegu, South Korea.
M AN U
6
SC
5
8
9
*Corresponding author:
Hyung-Kee Kim, M.D., Associate Professor
11
Division of Vascular Surgery, Department of Surgery
12
Kyungpook National University School of Medicine
13
130, Dongduk-ro, Jung-gu, Daegu, 700-721, South Korea
14
Phone: +82-53-420-5605
15
Fax: +82-53-421-0510
16
E-mail address: [email protected]
EP
AC C
17
TE D
10
18
Article type: Original Article
19
Running title: Reintervention after treatment of abdominal aortic aneurysm
20
Funding acknowledgement: None
21
Conflict of Interest: None
ACCEPTED MANUSCRIPT
EP
TE D
M AN U
SC
RI PT
Financial Disclosure: None
AC C
22
ACCEPTED MANUSCRIPT
Abstract
24
Background: We aim to determine the reintervention rate after open aortic aneurysm repair
25
(OAR) or endovascular aneurysm repair (EVAR) according to compliance or noncompliance
26
with the instructions for use (IFU) for commercial endovascular stent grafts.
27
Methods: After exclusion of those with a ruptured abdominal aortic aneurysm (AAA) and
28
isolated iliac artery aneurysm with or without a small AAA (diameter < 5cm), 240 patients
29
received OAR or EVAR for a nonruptured AAA between January 2006 and March 2016.
30
EVAR was performed from October 2009. Patients were divided into three groups: OAR (n =
31
146), IFU EVAR (n = 42), and non-IFU EVAR (n = 52). Reintervention was defined as graft-
32
related or laparotomy-related (with an abdominal incision after initial laparotomy) re-
33
operations either during the index admission period or later. Final endoleak after EVAR was
34
defined as persistent Type I or III endoleak before exiting operating room after various
35
procedures to eliminate the endoleak.
36
Results: There were two in-hospital deaths in the OAR group caused by reperfusion injury or
37
pancreatitis. There was no in-hospital mortality in the EVAR group. Final endoleak was more
38
common in non-IFU EVAR compared with IFU EVAR (17% vs 0%; P = 0.004). The mean
39
follow-up duration was 42.1 months, 25.3 months, and 25.0 months in the OAR, IFU EVAR,
40
and non-IFU EVAR groups, respectively. Respective reintervention-free survival (RFS) rates
41
at 1 and 3 years differed significantly by group: 97% and 95% in the OAR group, 100% and
42
96% in the IFU EVAR group, and 89% and 87% for non-IFU EVAR group (P = 0.043) with a
43
higher reintervention rate in the non-IFU EVAR than in the OAR group. There was no
44
significant difference in RFS rate between the OAR and IFU EVAR groups (P = 0.881).
45
Overall survival (OS) rates at 1 and 3 years, respectively, were 94% and 78% in the OAR
46
group, 90% and 86% in the IFU EVAR group, and 93% and 56% in the non-IFU EVAR
AC C
EP
TE D
M AN U
SC
RI PT
23
ACCEPTED MANUSCRIPT group (P = 0.098). There were no significant differences between the OAR and IFU EVAR
48
groups (P = 0.890).
49
Conclusions: In contrast to IFU EVAR group, the RFS and OS rates of non-IFU EVAR
50
group were lower than in the OAR group during mid-term follow-up. Final endoleak was
51
more common and reintervention was more commonly performed in the non-IFU group than
52
in the IFU group. Therefore, performing EVAR in non-IFU situations should be planned
53
carefully.
SC
RI PT
47
54
Key Words: Abdominal aortic aneurysm, Operative procedures, Endovascular procedures,
56
reoperation
AC C
EP
TE D
M AN U
55
ACCEPTED MANUSCRIPT
1. INTRODUCTION
58
Endovascular aneurysm repair (EVAR) for an abdominal aortic aneurysm (AAA) was first
59
described by Parodi et al. in 1991 and has been used commonly over the past two decades.1,2
60
Two meta-analyses analyzing randomized controlled trials (RCTs) that compared the
61
outcomes between open aortic aneurysm repair (OAR) and EVAR in cases of AAA have been
62
published.3,4 According to these, the 30-day mortality following OAR was higher than that of
63
EVAR. In contrast, more reinterventions tend to be conducted in cases of EVAR.3,4 However,
64
patients included in those trials were all surgically favorable and their anatomy conformed to
65
the instructions for use (IFU) of commercial endovascular stent grafts. There are also
66
discrepancies in the definition of “reintervention”, which has had various meanings among
67
trials. While laparotomy-related procedures and wound-related problems were included in the
68
OVER and ACE trials, graft-related complications were only considered as reinterventions in
69
the EVAR1 and DREAM trials.5-8 Presently, performing EVAR while not adhering to the IFU
70
for applying endovascular stent grafts is common. Therefore, the results of existing meta-
71
analyses do not adequately reflect “real-world” experience.
72
In addition, there is still controversy on whether application of the IFU for cases of EVAR
73
affects the outcomes. Generally, it is expected that reintervention after EVAR will be
74
increased in cases with EVAR not adhering to the IFU compared with those that adhere to the
75
IFU as time goes on. However, according to the recently published long-term follow-up data
76
between IFU-adherent and IFU-nonadherent EVAR, postoperative results suggest that there
77
are no significant differences in aneurysm sac size changes, overall and aneurysm-related
78
mortality, rates of endoleak or reintervention, and mid-term outcomes regardless of suitability
79
for the IFU.9-11 Furthermore, comparisons of reinterventions between cases of OAR and
80
EVAR outside the IFU have not yet been sufficiently reported.
AC C
EP
TE D
M AN U
SC
RI PT
57
ACCEPTED MANUSCRIPT Therefore, we aimed to review mid-term experiences at our center and to determine the
82
reintervention rates and overall survival after OAR and EVAR according to compliance with
83
the IFU of commercial endovascular stent grafts in patients with a nonruptured AAA to assess
84
the safety of EVAR.
RI PT
81
85
2. METHODS
87
2. 1 Data Sources and Variables
88
This study was approved ethically and supervised from a medical perspective through our
89
Institutional Review Board. From January 2006 to March 2016, 327 patients received repair
90
for AAA or iliac artery aneurysm in our medical center, a major tertiary hospital in South
91
Korea. After excluding cases of ruptured AAAs and iliac artery aneurysms with or without a
92
small AAA (diameter 3–5cm), a total of 240 patients with a nonruptured AAA larger than
93
5cm were eligible for this study. OAR was conducted in 146 patients and EVAR was
94
performed in 94. EVAR was performed from October 2009. Among patients receiving EVAR,
95
42 were anatomically compliant with the IFU and 52 patients were not, according to the
96
manufacturers’ guides. We have measured anatomical parameters of AAA for Endurant and
97
Endurant IIs (Medtronic Cardiovascular, Santa Rosa, CA, USA), or Excluder and C3 (WL
98
Gore and Associates, Flagstaff, AZ, USA) devices. These two devices usually have been used
99
in accordance with surgeon’s preference, familiarity, and patient’s anatomical suitability. Data
100
of patient and aneurysm characteristics were analyzed retrospectively, and intraoperative
101
findings during EVAR were collected prospectively. Reintervention and out-of-hospital
102
mortality were investigated using a chart review and telephone canvassing. Variables such as
103
age, gender, and comorbidities such as hypertension, diabetes mellitus, coronary artery
104
disease, congestive heart failure, arrhythmia, cerebral vascular disease, chronic obstructive
AC C
EP
TE D
M AN U
SC
86
ACCEPTED MANUSCRIPT pulmonary disease, renal insufficiency (serum creatinine > 1.5mg/dl), and chronic liver
106
disease were extracted as patients’ characteristics.
107
Our outcomes of interest were the characteristics of the patients and aneurysms, kind of
108
reintervention and reintervention-free survival, and overall survival in the OAR and EVAR
109
groups. These outcomes were also compared in respect to the suitability of AAA according to
110
anatomical compliance with the IFU (IFU EVAR and non-IFU EVAR groups). In addition,
111
we compared the findings of completion and final angiography among the EVAR group.
112
2.2 Diagnostic, Therapeutic, and Follow-up Protocols
113
Computed tomography (CT) images were processed and reconstructed into three-dimensional
114
images via AquarisNET software (TeraRecon Inc., San Mateo, CA, USA). Sac diameter, neck
115
length, proximal and distal neck diameter, neck shape, suprarenal and infrarenal angle, distal
116
aortic diameter, combined iliac artery aneurysm, and iliac tortuosity were examined in the
117
EVAR group. Sac diameter was calculated as the anteroposterior length to be compatible with
118
measurements made using duplex ultrasound (DUS). All values from CT images were
119
measured and classified by a vascular surgeon (H.K.) and an interventional radiologist (J.L.).
120
We considered an infrarenal fixation device as the primary choice. Seventy patients (74%) in
121
the EVAR group were treated with infrarenal fixation devices (Excluder or C3) and the rest of
122
the patients (n = 24, 26%) received suprarenal fixation devices (Endurant or Endurant IIs). In
123
the EVAR group, follow-up images were planned to be taken at 1 and 6months, 1year, and
124
annually thereafter after the EVAR, using CT or DUS. When sac enlargement with endoleak
125
or migration were suspected, they were checked with CT. DUS was preferred for patients
126
with renal impairment.
AC C
EP
TE D
M AN U
SC
RI PT
105
ACCEPTED MANUSCRIPT 2.3 Definitions
128
2.3.1 IFU or Non-IFU
129
Whether the morphology of an aneurysm was suitable for the IFU was determined according
130
to the recommendation of each manufacturer. When unfavorable neck anatomy including a
131
short neck, angulation, and conical neck or inadequate iliac features such as a short landing
132
zone or wide diameter that did not match each IFU were checked, the patient was categorized
133
as being in the non-IFU EVAR group.
134
2.3.2 Angulation
135
As widely accepted, we regarded it as infrarenal angulation when an angle between the axis
136
of aneurysm and the axis of infrarenal neck has a value exceeding 60 degrees and as
137
suprarenal angulation when an angle between the axis of infrarenal neck and the axis of
138
suprarenal aorta is exceeding 45 degrees (75 and 60 degrees, respectively in Endurant device
139
with neck length 15mm or more).
140
2.3.3 Final Endoleak
141
After completion of endograft placement and before the removal of the delivery system, a
142
completion angiogram was checked. In cases with Type I or III endoleaks, various additional
143
procedures were performed to eliminate the endoleak. Despite these additional procedures,
144
this was recorded as the final endoleak when any suspicious leakage of contrast medium
145
through the endograft was found.
146
2.3.4 Reintervention
147
We defined the meaning of “reintervention” as a graft-related or laparotomy-related re-
148
operation, and the procedure could be performed either during the index admission period or
149
at a later admission. We also defined laparotomy-related re-operation as redo operation that
150
was done with an abdominal incision after initial laparotomy.
AC C
EP
TE D
M AN U
SC
RI PT
127
ACCEPTED MANUSCRIPT 2.3.5 In-Hospital Mortality
152
If a patient died postoperatively during the index admission period, the case was considered
153
as in-hospital mortality, regardless of the length of hospital stay.
154
2.4 Data Analysis
155
Results were analyzed statistically using SPSS v. 20.0 for Windows (IBM Corp., Armonk, NY,
156
USA). Fourteen categorical variables were subjected to chi-squared analysis (if sample size
157
was adequate) or Fisher’s exact test (for smaller samples). For continuous variables with
158
normal distribution, the data are presented as mean and standard deviation; for variables with
159
non-normal distribution, the data are presented as median and interquartile range. A one-way
160
analysis of variance (ANOVA) followed by Tukey's post hoc test was used to analyze
161
differences of age between groups. For the comparison of characteristics of aneurysm
162
between IFU EVAR and non-IFU EVAR groups, student’s independent t test was applied for
163
comparison of means after confirming normality of distributions. Given the potential for
164
skew, group comparison of suprarenal and infrarenal angulation, and diameters of both CIA
165
relied on the Mann–Whitney nonparametric U test (also known as the Wilcoxon test for
166
independent measures). Kaplan–Meier plots were used to assess reintervention-free survival
167
and overall survival rates. Differences between the survival curves were evaluated for
168
significance by the log-rank test, and significance was assumed at P < 0.05.
SC
M AN U
TE D
EP
AC C
169
RI PT
151
170
3. RESULTS
171
3.1 Patients, Aneurysm Characteristics, and Device Selection
172
The patients’ characteristics are summarized in Table I, and aneurysm characteristics in the
173
EVAR group are summarized in Table II. Patients receiving OAR were significantly younger
174
than in both of the EVAR groups (P = 0.003). The proportion of female patients was higher in
175
the non-IFU EVAR group as 25% (13/52), compared with 14% (20/146) in the OAR group
ACCEPTED MANUSCRIPT and 7% (3/42) in the IFU EVAR group (P = 0.043). The rates of patients with preoperative
177
arrhythmia were higher in both of the EVAR groups compared with OAR group (IFU EVAR,
178
n = 12, 29%; non-IFU EVAR, n = 10, 20%; OAR, n = 11, 8%; P = 0.001) and a significantly
179
higher rate of CAD (n = 25, 48%) was noted in the non-IFU EVAR group (P = 0.001)
180
compared with the IFU EVAR and OAR groups.
181
A comparison of aneurysm characteristics between the two EVAR groups showed larger
182
suprarenal and infrarenal angulations in the non-IFU EVAR group (P < 0.001, Table II).
183
However, in regard to neck length, there were no significant intergroup differences, as 32.6
184
and 31.1mm, respectively (P = 0.504). In terms of maximal and distal aortic diameters,
185
aneurysms in the non-IFU EVAR group had larger values (58.4 and 29.4mm, respectively),
186
than those of IFU EVAR group (55.5 and 25.6mm; P = 0.079 and P = 0.031, respectively).
187
The right common iliac artery (CIA) in the non-IFU EVAR group was shorter than in the IFU
188
EVAR group (P = 0.031). As a cause of IFU unsuitability of 52 patients (Table II), a hostile
189
neck anatomy was found in 35 cases (67%) of the non-IFU EVAR group. Among these, the
190
rate of having an angulated neck was the highest at 82% (n = 29). An iliac cause of not
191
complying with the IFU accounted for 57% (30/52) of these cases, of which 14 (46%)
192
patients showed a short CIA and 10 (33%) showed aneurysms of the CIA, respectively.
193
Thirteen patients (25%) in the non-IFU EVAR group demonstrated both an unfavorable neck
194
and unfavorable iliac morphology.
195
3.2 Completion and Final Angiography Findings in the EVAR Group
196
On completion angiography, a Type Ia endoleak occurred in five (12%) patients in the IFU
197
EVAR group and in 12 (23%) of the non-IFU EVAR group (P = 0.162, Table III). Except for
198
one open conversion performed in the non-IFU EVAR group, endovascular means such as
199
additional balloon molding, Palmaz stents, and aortic cuffs were usually applied. There were
200
three (7%) and 11 (21%) cases of Type Ib endoleak in each group on completion angiography
AC C
EP
TE D
M AN U
SC
RI PT
176
ACCEPTED MANUSCRIPT (P = 0.058). Additional balloon molding was used in six patients (one in the IFU EVAR
202
group; five in the non-IFU EVAR group) and an additional iliac extender or cuff was used in
203
eight patients (two in the IFU EVAR group; six in the non-IFU EVAR group). Five cases of
204
Type III endoleak were detected in three of the IFU group and two of non-IFU EVAR groups
205
on completion angiography. Balloon moldings were applied in four cases. In one case of Type
206
III endoleak in the IFU EVAR group, the endoleak was persistent from the flow divider of the
207
stent graft; therefore, a double-barrel stent graft was used for sealing the endoleak.
208
After additional endovascular procedures, a final Type I endoleak was demonstrated in nine
209
(17%) patients of the non-IFU EVAR group (six Type Ia and three Type Ib) and no Type I
210
endoleak was found in the IFU EVAR group (P = 0.004). All of these nine patients were
211
closely followed with regular imaging work-up because the endoleak was small in all of them.
212
There was no Type III endoleak after additional procedures in both groups.
213
3.3 Type of Reintervention and Reintervention-free Survival (RFS)
214
In the OAR group, the mean follow-up duration was 42.1 months (range, 1–117) and nine
215
reinterventions were conducted in eight patients. Five cases of laparotomy-related
216
complication occurred in four patients and were repaired operatively, comprising three
217
incisional hernias in two patients, one intestinal obstruction, and one heterotopic ossification.
218
Four vascular complications occurred in four patients. During the index admission period,
219
one acute superficial femoral artery (SFA) occlusion and one graft occlusion developed
220
immediately after the operation. Femoral–popliteal artery bypass was conducted for the SFA
221
occlusion, and graft thrombectomy was conducted for the graft occlusion. One limb occlusion
222
occurred 12.6 months later and a femorofemoral bypass was performed. One internal iliac
223
artery aneurysm arose 6 years after initial operation and embolization was carried out.
224
In the IFU EVAR group, patients were followed up for an average of 25.3 months (range, 1–
225
87). During this period, three reinterventions were required for two patients. One patient
AC C
EP
TE D
M AN U
SC
RI PT
201
ACCEPTED MANUSCRIPT suspected of having a graft infection underwent debridement of the aortic sac and
227
omentopexy 15 months later. In another patient diagnosed with a Type III endoleak 3 years
228
after EVAR, relining with a new stent graft was conducted; however, a Type Ib endoleak
229
combined with rupture occurred 1 month later and an end-to-end anastomosis of the distal
230
stent graft to the CIA was implemented.
231
In the non-IFU EVAR group, the mean follow-up duration was 25 months (range, 1–74),
232
during which 11 reinterventions were carried out in seven patients, associated with 10
233
vascular complications and one general complication. Reinterventions were required for a
234
Type Ia endoleak in four cases, a Type Ib endoleak in three cases, a Type III endoleak in one
235
case, a Type II endoleak in one case, and limb occlusion in one case. For the Type Ia
236
endoleaks, one open conversion, two aortic cuff extensions, and one balloon molding were
237
used. In the three Type Ib endoleaks, stent graft extensions with or without embolization of
238
the internal iliac artery were performed. Open exploration and suture of the stent graft was
239
carried out for the one Type III endoleak and embolization of the inferior mesenteric artery
240
was performed for the sac expansion related to the Type II endoleak. In the patient with limb
241
occlusion, femorofemoral bypass was conducted. Among 9 patients with final Type I
242
endoleak in non-IFU EVAR group, the follow up assessment confirmed resolution of Type I
243
endoleak in 6 patients without any reinterventions, however, 3 patients required
244
reinterventions afterwards and all involved Type Ia endoleaks. As a nonvascular complication,
245
primary repair was conducted in a case of duodenal ulcer perforation during index admission
246
period.
247
Overall RFS rates at 1, 3, and 5 years, respectively, were 97, 95, and 95% in the OAR group,
248
100, 96, and 87% in the IFU EVAR group, and 89, 87, and 72% in the non-IFU EVAR group
249
(P = 0.043, Fig.1). Post-hoc comparisons revealed that the RFS rate in the non-IFU EVAR
250
group was significantly lower than in the OAR group (P = 0.020). However, there were no
AC C
EP
TE D
M AN U
SC
RI PT
226
ACCEPTED MANUSCRIPT significant differences between the OAR and IFU EVAR groups (P = 0.881), or between the
252
IFU EVAR and non-IFU EVAR groups (P = 0.128).
253
3.4 Overall Survival (OS) Rates
254
There was no in-hospital mortality in EVAR group. In the OAR group, 2/146 (1.4%) patients
255
died postoperatively during the index admission period (P = 0.521). Of these, one patient
256
suffered from severe pancreatitis and died of multisystem organ failure subsequent to septic
257
shock on postoperative day 76. The other patient with symptomatic AAA of 6.7cm in
258
diameter received an urgent operation. Subsequently, graft thrombosis with dissection at the
259
proximal clamping site occurred and this was detected during a reoperation. Despite
260
undergoing thrombectomy, the patient could not be recovered from reperfusion injury and
261
died 1 day after the reoperation.
262
OS rates at 1, 3, and 5 years, respectively, were 94, 78, and 71% in the OAR group, 90, 86,
263
and 77% in the IFU EVAR group, and 93, 56, and 45% in the non-IFU EVAR group (P =
264
0.098, Fig.2). Post-hoc comparisons revealed that the OS rate in the non-IFU EVAR group
265
was significantly lower than in the OAR group (P = 0.031). However, there were no
266
significant differences between the OAR and IFU EVAR group (P = 0.890), or between the
267
IFU EVAR and non-IFU EVAR groups (P = 0.207).
268
After the procedure, in OAR group 13 patients were lost to follow up and 39 patients expired,
269
of whom causes of death were unknown in 7 patients. Other causes of death included heart
270
disease (n=11; CAD, heart failure), malignancy (n=8), cerebral infarction (n=6), pulmonary
271
disease (n=4; COPD, pneumonia), cardiac tamponade due to ascending aortic rupture (n=1),
272
intestinal obstruction (n=1), and reperfusion injury (n=1). Meanwhile, no one was lost to
273
follow up in both of EVAR groups. Causes of death included malignancy (n=3), suicide (n=1),
274
heart failure (n=1), and unknown causes (n=2) in IFU group, while those included pulmonary
275
disease (n=4; pneumonia, COPD), heart failure (n=3), malignancy (n=2), acute myocardial
AC C
EP
TE D
M AN U
SC
RI PT
251
ACCEPTED MANUSCRIPT 276
infarction (n=1), acute renal failure (n=1), peritonitis (n=1), and unknown causes (n=3) in
277
non-IFU EVAR group. After exclusion of deaths of unknown cause, no aneurysm-related
278
death was identified in EVAR group.
RI PT
279
4. DISCUSSION
281
In contrast to the lower perioperative mortality after EVAR, the long-term results of EVAR
282
found in recent RCTs suggest that EVAR does impair survival and leads to higher
283
reintervention rates. This in turn leads to the convergence of survival rates to a similar extent
284
as those seen after open repair.6,8,12,13 However, there is controversy in regard to
285
reinterventions after OAR and EVAR. As described above, the definition of reintervention
286
has differed according to RCTs, and especially whether laparotomy- or wound-related
287
reinterventions after OAR are included. For example, in contrast to previously described
288
RCTs, Schermerhorn et al. reported a comparison of EVAR with OAR for an AAA in the
289
Medicare population, and demonstrated that late reinterventions related to AAA were more
290
common after EVAR but were balanced by an increase in laparotomy-related reinterventions
291
and hospitalizations after OAR.14,15 In our study, RFS and OS rates in non-IFU EVAR group
292
were significantly lower than that of OAR group. However, there were no differences in RFS
293
or OS rates between the OAR and IFU EVAR groups during a mid-term follow-up. In
294
addition, the patients in EVAR were older and had higher rate of comorbidities such as CAD,
295
CHF, and arrhythmia compared with the OAR group. Therefore, EVAR can be safely used for
296
the treatment of patients with unruptured AAA who are suitable for the IFU, although those
297
patients still need to be under close surveillance for detecting delayed endoleaks.
298
Currently, technological advances and accumulated experience have allowed an increasing
299
number of high-risk patients to receive EVAR. In fact, the number of patients who undergo
300
EVAR is on the increase even if the anatomy of the aneurysm is unsuitable for the IFU.
AC C
EP
TE D
M AN U
SC
280
ACCEPTED MANUSCRIPT However, the application of EVAR outside the IFU can cause a potential rupture of the AAA
302
following aneurysm sac enlargement.16 In our study, more reinterventions were conducted in
303
the non-IFU EVAR group than other groups. Final angiography revealed that there were 6
304
Type Ia endoleaks and 3 Type Ib endoleaks in the non-IFU EVAR group. Although the follow
305
up assessment confirmed resolution of Type I endoleak in 6 patients without any
306
reinterventions, 3 of 9 cases required reintervention afterwards and all involved Type Ia
307
endoleaks. In regard to overall survival, the survival rate in both the OAR and non-IFU
308
EVAR groups remained similar at about 90% up to 2 years after operation; however, it fell
309
sharply in the non-IFU EVAR group thereafter. It appears that this was caused by higher
310
proportions of elderly patients and comorbidities. In other words, the OS rate was
311
significantly lower in the non-IFU EVAR group compared with the OAR group after 2 years
312
had passed, and the reintervention rate was higher than in the OAR group during the follow-
313
up period. Therefore, patients treated with non-IFU EVAR should be under strict surveillance
314
for the development of complications.
315
Proximal neck features remain the most important factors determining patient suitability for
316
the IFU for EVAR.9 Although the effect of a short neck on the outcome is obvious, the role of
317
neck angulation is still disputable. As demonstrated in previous studies comparing hostile
318
neck anatomy with a more favorable one, a more angled aortic neck can induce more
319
complications of Type I endoleaks, migrations, adjunctive interventions, and sac
320
expansions.17-20 However, when studying the application of EVAR using the Endurant device
321
with about a 4-year follow-up, primary clinical success and freedom from neck-related
322
complications and reinterventions did not differ according to the extent of angulation.21 While
323
the neck lengths of patients in the above studies also differed between groups in addition to
324
angulation, our patients had similar neck lengths between groups but with more severe
325
angulation in the non-IFU EVAR group. This difference in anatomy appears to have
AC C
EP
TE D
M AN U
SC
RI PT
301
ACCEPTED MANUSCRIPT reinforced the effect of neck angulation on EVAR outcome. In this study, reinterventions
327
were performed in 3 cases treated with IFU EVAR and 11 cases with non-IFU EVAR. Among
328
them, there were no reinterventions caused by proximal attachment problem in the IFU
329
EVAR group, however, 4 (36%) cases of reintervention were caused by Type Ia endoleak in
330
the non-IFU EVAR group. In addition, irrespective of the reasons for reintervention, 5 of 6
331
patients undergoing graft-related reintervention in non-IFU EVAR group had angulated necks.
332
This suggests that neck angulation is an important obstacle to be overcome for better
333
outcomes after non-IFU EVAR.
334
Unlike western countries with frequent short aortic neck in non-IFU EVAR, our study shows
335
that aortic neck angulation acts as an important limitation regarding the IFU of endograft. It is
336
noteworthy that the proportion of female gender was significantly higher in non-IFU EVAR
337
group than other groups (non-IFU EVAR, 25%; OAR, 14%; IFU EVAR, 7%, P=0.043). It
338
turned out that 10 of 13 female patients in non-IFU EVAR group had angulated neck. This
339
finding is similar to the result of previous studies that female gender is less likely to meet the
340
IFU for EVAR owing to unfavorable neck anatomy or poor iliac access vessels.22-24 In
341
addition, more research in order to verify the anatomical variation by race is required. It
342
seems that the effect of race on the outcomes in EVAR has not been much investigated.
343
Existing RCTs usually have been conducted in the west and the study centers were almost
344
located in Europe or United states. Besides, in most RCTs, analysis according to race has not
345
been performed. Even in a few studies mentioning the categorization by race, the Caucasian
346
race forms more than 80% of patients.13 If the outcomes were compared by various races
347
consisting of an equivalent number of patients, more significant results in the relationship
348
between anatomical characteristics and race could be drawn, which may result in remarkable
349
progress in EVAR following accelerated development of devices compatible with a specific
350
race.
AC C
EP
TE D
M AN U
SC
RI PT
326
ACCEPTED MANUSCRIPT One of the limitations of this study is that relatively few patients were enrolled and followed
352
up during a short period of time, and the different average durations among groups. Moreover,
353
fewer patients were assigned to the EVAR procedure than in the OAR group. This must have
354
been caused by the late introduction of the EVAR procedure compared with OAR. Device
355
selection strongly reflected the surgeon’s preference. The selection of endograft was also
356
heavily weighted toward some specific devices. We have mainly used infrarenal fixation
357
devices (Excluder or C3) because of the well-known lower rate of limb occlusion and
358
recently reported additional merits of repositioning after deployment. However, it is known
359
that the use of a familiar device will enhance the outcome of any procedure. Thus, following
360
the surgeon’s preference was not problematic even though it might have influenced the
361
results. Even though the major factors creating a conflict in regard to the execution of EVAR
362
outside the IFU are long-term durability and safety, unfortunately, this study did not show
363
that how many patients among 12 demises of unknown cause exactly died of rupture- or
364
aneurysm-related complications. An underestimation of aneurysm related complication or
365
mortality may have occurred because of the lack of information about cause of death.
366
Whether performing EVAR in patients outside the IFU is ethical is a controvertible issue.
367
However, it is well known that there are no definite alternatives for patients who have
368
prohibitive risk for operation and concurrently do not satisfy the IFU of commercial EVAR
369
devices.25 Even though patients did not meet the IFU, we often decided to perform EVAR
370
instead of OAR on account of their poor general conditions. However, as in our series, non-
371
IFU EVAR has been known to be associated with higher rate of failed EVAR and
372
complications during follow-up. Therefore, risk of open surgery, risk and long-term
373
complications after EVAR in unsuitable anatomy, patients’ preferences, and life expectancy
374
should be considered and discussed thoroughly. This is not a randomized, prospective study,
375
but a retrospective design in a single center, so the results should be interpreted carefully and
AC C
EP
TE D
M AN U
SC
RI PT
351
ACCEPTED MANUSCRIPT 376
without making generalizations. Nonetheless, there is importance in mirroring real-world
377
surgical practice without manipulating it artificially.
378
5. CONCLUSIONS
380
In the IFU EVAR group, the RFS and OS rates were similar to the OAR group during mid-
381
term follow-up. Meanwhile, in the non-IFU group, the RFS and OS rates were lower than in
382
the OAR group and final endoleak was more common than in the IFU EVAR group.
383
Although not statistically significant, reintervention was more commonly performed in the
384
non-IFU EVAR than in the IFU EVAR group. Therefore, performing EVAR in non-IFU
385
situations should be planned carefully and under strict surveillance for the development of
386
complications.
387
TE D EP
389
6. ACKNOWLEDGEMENT
AC C
388
M AN U
SC
RI PT
379
ACCEPTED MANUSCRIPT 390
REFERENCES
391
1.
abdominal aortic aneurysms. Ann Vasc Surg 1991;5:491-9. 2.
nonagenarians: a systematic review. Ann Vasc Surg 2015;29:385-91.
394 395
Wigley J, Shantikumar S, Hameed W, et al. Endovascular aneurysm repair in
RI PT
392 393
Parodi JC, Palmaz JC, Barone HD. Transfemoral intraluminal graft implantation for
3.
Dangas G, O'Connor D, Firwana B, et al. Open versus endovascular stent graft repair of abdominal aortic aneurysms: a meta-analysis of randomized trials. JACC
397
Cardiovasc Interv 2012;5:1071-80. 4.
elective abdominal aortic aneurysm repair. J Vasc Surg 2013;57:1676-83, 83 e1.
399 400
Qadura M, Pervaiz F, Harlock JA, et al. Mortality and reintervention following
M AN U
398
SC
396
5.
Becquemin JP. The ACE trial: a randomized comparison of open versus endovascular repair in good risk patients with abdominal aortic aneurysm. J Vasc Surg
402
2009;50:222-4; discussion 4. 6.
repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1881-9.
404 405
De Bruin JL, Baas AF, Buth J, et al. Long-term outcome of open or endovascular
7.
Lederle FA, Freischlag JA, Kyriakides TC, et al. Outcomes following endovascular vs
EP
403
TE D
401
open repair of abdominal aortic aneurysm: a randomized trial. JAMA 2009;302:1535-
407
42.
408
8.
9.
Lee JT, Ullery BW, Zarins CK, et al. EVAR deployment in anatomically challenging
necks outside the IFU. Eur J Vasc Endovasc Surg 2013;46:65-73.
411 412
United Kingdom ETI, Greenhalgh RM, Brown LC, et al. Endovascular versus open
repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1863-71.
409 410
AC C
406
10.
Walker J, Tucker LY, Goodney P, et al. Adherence to endovascular aortic aneurysm
413
repair device instructions for use guidelines has no impact on outcomes. J Vasc Surg
414
2015;61:1151-9.
ACCEPTED MANUSCRIPT 415
11.
Zandvoort HJ, Goncalves FB, Verhagen HJ, et al. Results of endovascular repair of
416
infrarenal aortic aneurysms using the Endurant stent graft. J Vasc Surg 2014;59:1195-
417
202. 12.
Becquemin JP, Pillet JC, Lescalie F, et al. A randomized controlled trial of
RI PT
418 419
endovascular aneurysm repair versus open surgery for abdominal aortic aneurysms in
420
low- to moderate-risk patients. J Vasc Surg 2011;53:1167-73 e1. 13.
Lederle FA, Freischlag JA, Kyriakides TC, et al. Long-term comparison of
SC
421
endovascular and open repair of abdominal aortic aneurysm. N Engl J Med
423
2012;367:1988-97.
424
14.
M AN U
422
Schermerhorn ML, O'Malley AJ, Jhaveri A, et al. Endovascular vs. open repair of
425
abdominal aortic aneurysms in the Medicare population. N Engl J Med 2008;358:464-
426
74.
Aortic Aneurysm in the Medicare Population. N Engl J Med 2015;373:328-38.
428
16.
Endovascular Abdominal Aortic Aneurysm Repair. Ann Vasc Surg 2016;31:229-38.
430 431
17.
neck anatomy. J Vasc Surg 2013;57:527-38.
433
18.
Hobo R, Kievit J, Leurs LJ, et al. Influence of severe infrarenal aortic neck angulation on complications at the proximal neck following endovascular AAA repair: a
435
EUROSTAR study. J Endovasc Ther 2007;14:1-11.
436 437
Antoniou GA, Georgiadis GS, Antoniou SA, et al. A meta-analysis of outcomes of endovascular abdominal aortic aneurysm repair in patients with hostile and friendly
432
434
Dingemans SA, Jonker FH, Moll FL, et al. Aneurysm Sac Enlargement after
EP
429
Schermerhorn ML, Buck DB, O'Malley AJ, et al. Long-Term Outcomes of Abdominal
TE D
15.
AC C
427
19.
Malas MB, Jordan WD, Cooper MA, et al. Performance of the Aorfix endograft in
438
severely angulated proximal necks in the PYTHAGORAS United States clinical trial.
439
J Vasc Surg 2015;62:1108-17.
ACCEPTED MANUSCRIPT 440
20.
patients with hostile neck anatomy. Eur J Vasc Endovasc Surg 2012;44:556-61.
441 442
Stather PW, Sayers RD, Cheah A, et al. Outcomes of endovascular aneurysm repair in
21.
Oliveira NF, Bastos Goncalves FM, de Vries JP, et al. Mid-Term Results of EVAR in Severe Proximal Aneurysm Neck Angulation. Eur J Vasc Endovasc Surg 2015;49:19-
444
27.
445
22.
RI PT
443
Sweet MP, Fillinger MF, Morrison TM, et al. The influence of gender and aortic aneurysm size on eligibility for endovascular abdominal aortic aneurysm repair. J
447
Vasc Surg 2011;54:931-7. 23.
Dubois L, Novick TV, Harris JR, et al. Outcomes after endovascular abdominal aortic
M AN U
448
SC
446
449
aneurysm repair are equivalent between genders despite anatomic differences in
450
women. J Vasc Surg 2013;57:382-9 e1.
451
24.
Hultgren R, Vishnevskaya L, Wahlgren CM. Women with abdominal aortic aneurysms have more extensive aortic neck pathology. Ann Vasc Surg 2013;27:547-
453
52.
456 457
Rutherford RB. Management of abdominal aortic aneurysms: which risk factors play a role in decision-making? Semin Vasc Surg 2008;21:124-31.
EP
455
25.
AC C
454
TE D
452
ACCEPTED MANUSCRIPT 458
Table I. Characteristics of patients with open aneurism repair (OAR) and endovascular
459
aneurysm repair (EVAR) IFU EVAR
Non-IFU
(n = 146)
(n = 42)
EVAR
P
RI PT
OAR
(n = 52) 69.4 ± 7.8
72.2 ± 7.3
73.2 ± 6.7
0.003
Female
20 (14%)
3 (7%)
13(25%)
0.043
HTN
96 (66%)
24 (57%)
DM
16 (11%)
CAD
31 (21%)
CHF
7 (5%)
Arrhythmia
11 (8%)
CVD
18 (12%)
COPD Renal
0.889
13 (31%)
25 (48%)
0.001
5 (12%)
7 (14%)
0.058
12 (29%)
10 (20%)
0.001
2 (5%)
9 (17%)
0.177
22 (15%)
8 (19%)
13 (25%)
0.270
13 (9%)
4 (10%)
10 (19%)
0.120
2 (5%)
2 (4%)
1.000
TE D
7 (14%)
6 (4%)
AC C
CLD
0.453
5 (12%)
EP
insufficiencya
36 (69%)
M AN U
Comorbidities:
SC
Age, years
460
IFU, instructions for use; HTN, hypertension; DM, diabetes mellitus; CAD, coronary artery
461
disease; CHF, congestive heart failure; CVD, cerebrovascular disease; COPD, chronic
462
obstructive pulmonary disease; CLD, chronic liver disease.
463
464 465
a
Renal insufficiency was defined as preoperative serum creatinine level of 1.5 mg/dL or
higher.
ACCEPTED MANUSCRIPT 466
Table II. Aneurysm characteristics in endovascular aneurysm repair (EVAR) groups treated
467
according to the instructions for use (IFU EVAR), not according to the IFU (non-IFU EVAR),
468
and anatomic causes of unsuitability for the IFU in the non-IFU EVAR group. Non-IFU EVAR
(n = 42)
(n = 52)
Characteristics of aneurysms 14.9 (8.8-31.6)
39.1 (21.3-53.0)
0.000
Infrarenal angulation (°) (IQR)
31.2 (26.3-42.7)
67.0 (35.9-77.4)
0.000
Neck length (mm) (±SD)
32.6 (±10.5)
31.1 (±10.9)
0.504
Neck diameter (mm) (±SD)
21.1 (±2.4)
20.7 (±2.4)
0.457
AAA maximum diameter (mm) (±SD)
55.5 (±8.4)
58.4 (±7.2)
0.079
Distal aortic diameter (mm) (±SD)
25.6 (±7.1)
29.5 (±9.3)
0.031
13.9 (12.2-15.7)
14.7 (11.0-18.4)
0.525
14.2 (12.3-15.9)
14.2 (11.7-16.7)
0.981
CIA length, R. (mm) (±SD)
42.6 (±17.7)
34.9 (±16.4)
0.031
CIA length, L. (mm) (±SD)
45.0 (±19.5)
39.9 (±17.3)
0.184
Iliac tortuosity index, right side (±SD)
1.36 (±0.13)
1.35 (±0.12)
0.549
Iliac tortuosity index, left side (±SD)
1.34 (±0.14)
1.35 (±0.13)
0.713
AC C
EP
TE D
CIA diameter, L. (mm) (IQR)
M AN U
CIA diameter, R. (mm) (IQR)
SC
Suprarenal angulation (°) (IQR)
P
RI PT
IFU EVAR
Anatomic causes of IFU unsuitability Neck causea
35 (67%)
Angulated neck
29 (82%)
Iliac causeb
30 (57%)
Short CIA
14 (46%)
Aneurysm of CIA or EIA
10 (33%)
Infrarenal fixation device
34 (81%)
36 (69%)
ACCEPTED MANUSCRIPT Suprarenal fixation device
8 (19%)
16 (31%)
469
IQR, interquartile range; SD, standard deviation; AAA, abdominal aortic aneurysm; CIA,
470
common iliac artery; EIA, external iliac artery.
473 474
“Neck causes” in this row comprised a short neck (n = 4) and conical neck (n = 2), not
including an angulated neck. b
RI PT
472
a
“Iliac causes” in this row comprised a small EIA (n = 3), occlusion of the CIA or EIA (n =
2), and a narrow aortic bifurcation (n = 1).
SC
471
AC C
EP
TE D
M AN U
475
ACCEPTED MANUSCRIPT 476
Table III. Findings of completion angiography and subsequent treatments in endovascular
477
aneurysm repair groups treated according to the instructions for use (IFU EVAR), or not
478
according to the IFU (non-IFU EVAR).
(n = 42)
(n = 52)
5 (12%)
12 (23%)
Open conversion
1
Palmaz stent
1 2
5
Balloon molding
3
5
Type Ib endoleak
3 (7%)
Additional extender or cuff 2 Balloon molding
3 (7%)
Balloon molding Kissing stent graft
AC C
Type Ia Type Ib 479 480 481 482 483 484
2
0.162
11 (21%)
0.058
6
5
2 (4%)
0.653
2
1
0 (0%)
EP
Final Type I endoleak
TE D
Type III endoleak
1
M AN U
Aortic cuff
P
RI PT
Non-IFU EVAR
SC
Type Ia endoleak
IFU EVAR
9 (17%) 6 3
0.004
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
486
Fig. 1. Kaplan–Meier plots of reintervention-free survival rates in the OAR, IFU EVAR, and
487
non-IFU EVAR groups.
488
Fig. 2. Kaplan–Meier plots of overall survival rates in the OAR, IFU EVAR, and non-IFU
489
EVAR groups.
RI PT
485
AC C
EP
TE D
M AN U
SC
490
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0890-5096(16)30698-7
DOI:
10.1016/j.avsg.2017.03.168
Reference:
AVSG 3278
To appear in:
Annals of Vascular Surgery
Received Date: 21 August 2016 Revised Date:
2 February 2017
Accepted Date: 15 March 2017
Please cite this article as: Hwang D, Park S, Kim HK, Lee JM, Huh S, Reintervention Rate after Open Surgery and Endovascular Repair for Nonruptured Abdominal Aortic Aneurysms, Annals of Vascular Surgery (2017), doi: 10.1016/j.avsg.2017.03.168. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1
Reintervention Rate after Open Surgery and Endovascular Repair for
2
Nonruptured Abdominal Aortic Aneurysms
4
RI PT
3
Deokbi Hwanga, Sujin Parka, Hyung-Kee Kima, Jong-Min Leeb, and Seung Huha
7
a
Division of Vascular Surgery, Department of Surgery, bDepartment of Radiology, Kyungpook National University School of Medicine, Daegu, South Korea.
M AN U
6
SC
5
8
9
*Corresponding author:
Hyung-Kee Kim, M.D., Associate Professor
11
Division of Vascular Surgery, Department of Surgery
12
Kyungpook National University School of Medicine
13
130, Dongduk-ro, Jung-gu, Daegu, 700-721, South Korea
14
Phone: +82-53-420-5605
15
Fax: +82-53-421-0510
16
E-mail address: [email protected]
EP
AC C
17
TE D
10
18
Article type: Original Article
19
Running title: Reintervention after treatment of abdominal aortic aneurysm
20
Funding acknowledgement: None
21
Conflict of Interest: None
ACCEPTED MANUSCRIPT
EP
TE D
M AN U
SC
RI PT
Financial Disclosure: None
AC C
22
ACCEPTED MANUSCRIPT
Abstract
24
Background: We aim to determine the reintervention rate after open aortic aneurysm repair
25
(OAR) or endovascular aneurysm repair (EVAR) according to compliance or noncompliance
26
with the instructions for use (IFU) for commercial endovascular stent grafts.
27
Methods: After exclusion of those with a ruptured abdominal aortic aneurysm (AAA) and
28
isolated iliac artery aneurysm with or without a small AAA (diameter < 5cm), 240 patients
29
received OAR or EVAR for a nonruptured AAA between January 2006 and March 2016.
30
EVAR was performed from October 2009. Patients were divided into three groups: OAR (n =
31
146), IFU EVAR (n = 42), and non-IFU EVAR (n = 52). Reintervention was defined as graft-
32
related or laparotomy-related (with an abdominal incision after initial laparotomy) re-
33
operations either during the index admission period or later. Final endoleak after EVAR was
34
defined as persistent Type I or III endoleak before exiting operating room after various
35
procedures to eliminate the endoleak.
36
Results: There were two in-hospital deaths in the OAR group caused by reperfusion injury or
37
pancreatitis. There was no in-hospital mortality in the EVAR group. Final endoleak was more
38
common in non-IFU EVAR compared with IFU EVAR (17% vs 0%; P = 0.004). The mean
39
follow-up duration was 42.1 months, 25.3 months, and 25.0 months in the OAR, IFU EVAR,
40
and non-IFU EVAR groups, respectively. Respective reintervention-free survival (RFS) rates
41
at 1 and 3 years differed significantly by group: 97% and 95% in the OAR group, 100% and
42
96% in the IFU EVAR group, and 89% and 87% for non-IFU EVAR group (P = 0.043) with a
43
higher reintervention rate in the non-IFU EVAR than in the OAR group. There was no
44
significant difference in RFS rate between the OAR and IFU EVAR groups (P = 0.881).
45
Overall survival (OS) rates at 1 and 3 years, respectively, were 94% and 78% in the OAR
46
group, 90% and 86% in the IFU EVAR group, and 93% and 56% in the non-IFU EVAR
AC C
EP
TE D
M AN U
SC
RI PT
23
ACCEPTED MANUSCRIPT group (P = 0.098). There were no significant differences between the OAR and IFU EVAR
48
groups (P = 0.890).
49
Conclusions: In contrast to IFU EVAR group, the RFS and OS rates of non-IFU EVAR
50
group were lower than in the OAR group during mid-term follow-up. Final endoleak was
51
more common and reintervention was more commonly performed in the non-IFU group than
52
in the IFU group. Therefore, performing EVAR in non-IFU situations should be planned
53
carefully.
SC
RI PT
47
54
Key Words: Abdominal aortic aneurysm, Operative procedures, Endovascular procedures,
56
reoperation
AC C
EP
TE D
M AN U
55
ACCEPTED MANUSCRIPT
1. INTRODUCTION
58
Endovascular aneurysm repair (EVAR) for an abdominal aortic aneurysm (AAA) was first
59
described by Parodi et al. in 1991 and has been used commonly over the past two decades.1,2
60
Two meta-analyses analyzing randomized controlled trials (RCTs) that compared the
61
outcomes between open aortic aneurysm repair (OAR) and EVAR in cases of AAA have been
62
published.3,4 According to these, the 30-day mortality following OAR was higher than that of
63
EVAR. In contrast, more reinterventions tend to be conducted in cases of EVAR.3,4 However,
64
patients included in those trials were all surgically favorable and their anatomy conformed to
65
the instructions for use (IFU) of commercial endovascular stent grafts. There are also
66
discrepancies in the definition of “reintervention”, which has had various meanings among
67
trials. While laparotomy-related procedures and wound-related problems were included in the
68
OVER and ACE trials, graft-related complications were only considered as reinterventions in
69
the EVAR1 and DREAM trials.5-8 Presently, performing EVAR while not adhering to the IFU
70
for applying endovascular stent grafts is common. Therefore, the results of existing meta-
71
analyses do not adequately reflect “real-world” experience.
72
In addition, there is still controversy on whether application of the IFU for cases of EVAR
73
affects the outcomes. Generally, it is expected that reintervention after EVAR will be
74
increased in cases with EVAR not adhering to the IFU compared with those that adhere to the
75
IFU as time goes on. However, according to the recently published long-term follow-up data
76
between IFU-adherent and IFU-nonadherent EVAR, postoperative results suggest that there
77
are no significant differences in aneurysm sac size changes, overall and aneurysm-related
78
mortality, rates of endoleak or reintervention, and mid-term outcomes regardless of suitability
79
for the IFU.9-11 Furthermore, comparisons of reinterventions between cases of OAR and
80
EVAR outside the IFU have not yet been sufficiently reported.
AC C
EP
TE D
M AN U
SC
RI PT
57
ACCEPTED MANUSCRIPT Therefore, we aimed to review mid-term experiences at our center and to determine the
82
reintervention rates and overall survival after OAR and EVAR according to compliance with
83
the IFU of commercial endovascular stent grafts in patients with a nonruptured AAA to assess
84
the safety of EVAR.
RI PT
81
85
2. METHODS
87
2. 1 Data Sources and Variables
88
This study was approved ethically and supervised from a medical perspective through our
89
Institutional Review Board. From January 2006 to March 2016, 327 patients received repair
90
for AAA or iliac artery aneurysm in our medical center, a major tertiary hospital in South
91
Korea. After excluding cases of ruptured AAAs and iliac artery aneurysms with or without a
92
small AAA (diameter 3–5cm), a total of 240 patients with a nonruptured AAA larger than
93
5cm were eligible for this study. OAR was conducted in 146 patients and EVAR was
94
performed in 94. EVAR was performed from October 2009. Among patients receiving EVAR,
95
42 were anatomically compliant with the IFU and 52 patients were not, according to the
96
manufacturers’ guides. We have measured anatomical parameters of AAA for Endurant and
97
Endurant IIs (Medtronic Cardiovascular, Santa Rosa, CA, USA), or Excluder and C3 (WL
98
Gore and Associates, Flagstaff, AZ, USA) devices. These two devices usually have been used
99
in accordance with surgeon’s preference, familiarity, and patient’s anatomical suitability. Data
100
of patient and aneurysm characteristics were analyzed retrospectively, and intraoperative
101
findings during EVAR were collected prospectively. Reintervention and out-of-hospital
102
mortality were investigated using a chart review and telephone canvassing. Variables such as
103
age, gender, and comorbidities such as hypertension, diabetes mellitus, coronary artery
104
disease, congestive heart failure, arrhythmia, cerebral vascular disease, chronic obstructive
AC C
EP
TE D
M AN U
SC
86
ACCEPTED MANUSCRIPT pulmonary disease, renal insufficiency (serum creatinine > 1.5mg/dl), and chronic liver
106
disease were extracted as patients’ characteristics.
107
Our outcomes of interest were the characteristics of the patients and aneurysms, kind of
108
reintervention and reintervention-free survival, and overall survival in the OAR and EVAR
109
groups. These outcomes were also compared in respect to the suitability of AAA according to
110
anatomical compliance with the IFU (IFU EVAR and non-IFU EVAR groups). In addition,
111
we compared the findings of completion and final angiography among the EVAR group.
112
2.2 Diagnostic, Therapeutic, and Follow-up Protocols
113
Computed tomography (CT) images were processed and reconstructed into three-dimensional
114
images via AquarisNET software (TeraRecon Inc., San Mateo, CA, USA). Sac diameter, neck
115
length, proximal and distal neck diameter, neck shape, suprarenal and infrarenal angle, distal
116
aortic diameter, combined iliac artery aneurysm, and iliac tortuosity were examined in the
117
EVAR group. Sac diameter was calculated as the anteroposterior length to be compatible with
118
measurements made using duplex ultrasound (DUS). All values from CT images were
119
measured and classified by a vascular surgeon (H.K.) and an interventional radiologist (J.L.).
120
We considered an infrarenal fixation device as the primary choice. Seventy patients (74%) in
121
the EVAR group were treated with infrarenal fixation devices (Excluder or C3) and the rest of
122
the patients (n = 24, 26%) received suprarenal fixation devices (Endurant or Endurant IIs). In
123
the EVAR group, follow-up images were planned to be taken at 1 and 6months, 1year, and
124
annually thereafter after the EVAR, using CT or DUS. When sac enlargement with endoleak
125
or migration were suspected, they were checked with CT. DUS was preferred for patients
126
with renal impairment.
AC C
EP
TE D
M AN U
SC
RI PT
105
ACCEPTED MANUSCRIPT 2.3 Definitions
128
2.3.1 IFU or Non-IFU
129
Whether the morphology of an aneurysm was suitable for the IFU was determined according
130
to the recommendation of each manufacturer. When unfavorable neck anatomy including a
131
short neck, angulation, and conical neck or inadequate iliac features such as a short landing
132
zone or wide diameter that did not match each IFU were checked, the patient was categorized
133
as being in the non-IFU EVAR group.
134
2.3.2 Angulation
135
As widely accepted, we regarded it as infrarenal angulation when an angle between the axis
136
of aneurysm and the axis of infrarenal neck has a value exceeding 60 degrees and as
137
suprarenal angulation when an angle between the axis of infrarenal neck and the axis of
138
suprarenal aorta is exceeding 45 degrees (75 and 60 degrees, respectively in Endurant device
139
with neck length 15mm or more).
140
2.3.3 Final Endoleak
141
After completion of endograft placement and before the removal of the delivery system, a
142
completion angiogram was checked. In cases with Type I or III endoleaks, various additional
143
procedures were performed to eliminate the endoleak. Despite these additional procedures,
144
this was recorded as the final endoleak when any suspicious leakage of contrast medium
145
through the endograft was found.
146
2.3.4 Reintervention
147
We defined the meaning of “reintervention” as a graft-related or laparotomy-related re-
148
operation, and the procedure could be performed either during the index admission period or
149
at a later admission. We also defined laparotomy-related re-operation as redo operation that
150
was done with an abdominal incision after initial laparotomy.
AC C
EP
TE D
M AN U
SC
RI PT
127
ACCEPTED MANUSCRIPT 2.3.5 In-Hospital Mortality
152
If a patient died postoperatively during the index admission period, the case was considered
153
as in-hospital mortality, regardless of the length of hospital stay.
154
2.4 Data Analysis
155
Results were analyzed statistically using SPSS v. 20.0 for Windows (IBM Corp., Armonk, NY,
156
USA). Fourteen categorical variables were subjected to chi-squared analysis (if sample size
157
was adequate) or Fisher’s exact test (for smaller samples). For continuous variables with
158
normal distribution, the data are presented as mean and standard deviation; for variables with
159
non-normal distribution, the data are presented as median and interquartile range. A one-way
160
analysis of variance (ANOVA) followed by Tukey's post hoc test was used to analyze
161
differences of age between groups. For the comparison of characteristics of aneurysm
162
between IFU EVAR and non-IFU EVAR groups, student’s independent t test was applied for
163
comparison of means after confirming normality of distributions. Given the potential for
164
skew, group comparison of suprarenal and infrarenal angulation, and diameters of both CIA
165
relied on the Mann–Whitney nonparametric U test (also known as the Wilcoxon test for
166
independent measures). Kaplan–Meier plots were used to assess reintervention-free survival
167
and overall survival rates. Differences between the survival curves were evaluated for
168
significance by the log-rank test, and significance was assumed at P < 0.05.
SC
M AN U
TE D
EP
AC C
169
RI PT
151
170
3. RESULTS
171
3.1 Patients, Aneurysm Characteristics, and Device Selection
172
The patients’ characteristics are summarized in Table I, and aneurysm characteristics in the
173
EVAR group are summarized in Table II. Patients receiving OAR were significantly younger
174
than in both of the EVAR groups (P = 0.003). The proportion of female patients was higher in
175
the non-IFU EVAR group as 25% (13/52), compared with 14% (20/146) in the OAR group
ACCEPTED MANUSCRIPT and 7% (3/42) in the IFU EVAR group (P = 0.043). The rates of patients with preoperative
177
arrhythmia were higher in both of the EVAR groups compared with OAR group (IFU EVAR,
178
n = 12, 29%; non-IFU EVAR, n = 10, 20%; OAR, n = 11, 8%; P = 0.001) and a significantly
179
higher rate of CAD (n = 25, 48%) was noted in the non-IFU EVAR group (P = 0.001)
180
compared with the IFU EVAR and OAR groups.
181
A comparison of aneurysm characteristics between the two EVAR groups showed larger
182
suprarenal and infrarenal angulations in the non-IFU EVAR group (P < 0.001, Table II).
183
However, in regard to neck length, there were no significant intergroup differences, as 32.6
184
and 31.1mm, respectively (P = 0.504). In terms of maximal and distal aortic diameters,
185
aneurysms in the non-IFU EVAR group had larger values (58.4 and 29.4mm, respectively),
186
than those of IFU EVAR group (55.5 and 25.6mm; P = 0.079 and P = 0.031, respectively).
187
The right common iliac artery (CIA) in the non-IFU EVAR group was shorter than in the IFU
188
EVAR group (P = 0.031). As a cause of IFU unsuitability of 52 patients (Table II), a hostile
189
neck anatomy was found in 35 cases (67%) of the non-IFU EVAR group. Among these, the
190
rate of having an angulated neck was the highest at 82% (n = 29). An iliac cause of not
191
complying with the IFU accounted for 57% (30/52) of these cases, of which 14 (46%)
192
patients showed a short CIA and 10 (33%) showed aneurysms of the CIA, respectively.
193
Thirteen patients (25%) in the non-IFU EVAR group demonstrated both an unfavorable neck
194
and unfavorable iliac morphology.
195
3.2 Completion and Final Angiography Findings in the EVAR Group
196
On completion angiography, a Type Ia endoleak occurred in five (12%) patients in the IFU
197
EVAR group and in 12 (23%) of the non-IFU EVAR group (P = 0.162, Table III). Except for
198
one open conversion performed in the non-IFU EVAR group, endovascular means such as
199
additional balloon molding, Palmaz stents, and aortic cuffs were usually applied. There were
200
three (7%) and 11 (21%) cases of Type Ib endoleak in each group on completion angiography
AC C
EP
TE D
M AN U
SC
RI PT
176
ACCEPTED MANUSCRIPT (P = 0.058). Additional balloon molding was used in six patients (one in the IFU EVAR
202
group; five in the non-IFU EVAR group) and an additional iliac extender or cuff was used in
203
eight patients (two in the IFU EVAR group; six in the non-IFU EVAR group). Five cases of
204
Type III endoleak were detected in three of the IFU group and two of non-IFU EVAR groups
205
on completion angiography. Balloon moldings were applied in four cases. In one case of Type
206
III endoleak in the IFU EVAR group, the endoleak was persistent from the flow divider of the
207
stent graft; therefore, a double-barrel stent graft was used for sealing the endoleak.
208
After additional endovascular procedures, a final Type I endoleak was demonstrated in nine
209
(17%) patients of the non-IFU EVAR group (six Type Ia and three Type Ib) and no Type I
210
endoleak was found in the IFU EVAR group (P = 0.004). All of these nine patients were
211
closely followed with regular imaging work-up because the endoleak was small in all of them.
212
There was no Type III endoleak after additional procedures in both groups.
213
3.3 Type of Reintervention and Reintervention-free Survival (RFS)
214
In the OAR group, the mean follow-up duration was 42.1 months (range, 1–117) and nine
215
reinterventions were conducted in eight patients. Five cases of laparotomy-related
216
complication occurred in four patients and were repaired operatively, comprising three
217
incisional hernias in two patients, one intestinal obstruction, and one heterotopic ossification.
218
Four vascular complications occurred in four patients. During the index admission period,
219
one acute superficial femoral artery (SFA) occlusion and one graft occlusion developed
220
immediately after the operation. Femoral–popliteal artery bypass was conducted for the SFA
221
occlusion, and graft thrombectomy was conducted for the graft occlusion. One limb occlusion
222
occurred 12.6 months later and a femorofemoral bypass was performed. One internal iliac
223
artery aneurysm arose 6 years after initial operation and embolization was carried out.
224
In the IFU EVAR group, patients were followed up for an average of 25.3 months (range, 1–
225
87). During this period, three reinterventions were required for two patients. One patient
AC C
EP
TE D
M AN U
SC
RI PT
201
ACCEPTED MANUSCRIPT suspected of having a graft infection underwent debridement of the aortic sac and
227
omentopexy 15 months later. In another patient diagnosed with a Type III endoleak 3 years
228
after EVAR, relining with a new stent graft was conducted; however, a Type Ib endoleak
229
combined with rupture occurred 1 month later and an end-to-end anastomosis of the distal
230
stent graft to the CIA was implemented.
231
In the non-IFU EVAR group, the mean follow-up duration was 25 months (range, 1–74),
232
during which 11 reinterventions were carried out in seven patients, associated with 10
233
vascular complications and one general complication. Reinterventions were required for a
234
Type Ia endoleak in four cases, a Type Ib endoleak in three cases, a Type III endoleak in one
235
case, a Type II endoleak in one case, and limb occlusion in one case. For the Type Ia
236
endoleaks, one open conversion, two aortic cuff extensions, and one balloon molding were
237
used. In the three Type Ib endoleaks, stent graft extensions with or without embolization of
238
the internal iliac artery were performed. Open exploration and suture of the stent graft was
239
carried out for the one Type III endoleak and embolization of the inferior mesenteric artery
240
was performed for the sac expansion related to the Type II endoleak. In the patient with limb
241
occlusion, femorofemoral bypass was conducted. Among 9 patients with final Type I
242
endoleak in non-IFU EVAR group, the follow up assessment confirmed resolution of Type I
243
endoleak in 6 patients without any reinterventions, however, 3 patients required
244
reinterventions afterwards and all involved Type Ia endoleaks. As a nonvascular complication,
245
primary repair was conducted in a case of duodenal ulcer perforation during index admission
246
period.
247
Overall RFS rates at 1, 3, and 5 years, respectively, were 97, 95, and 95% in the OAR group,
248
100, 96, and 87% in the IFU EVAR group, and 89, 87, and 72% in the non-IFU EVAR group
249
(P = 0.043, Fig.1). Post-hoc comparisons revealed that the RFS rate in the non-IFU EVAR
250
group was significantly lower than in the OAR group (P = 0.020). However, there were no
AC C
EP
TE D
M AN U
SC
RI PT
226
ACCEPTED MANUSCRIPT significant differences between the OAR and IFU EVAR groups (P = 0.881), or between the
252
IFU EVAR and non-IFU EVAR groups (P = 0.128).
253
3.4 Overall Survival (OS) Rates
254
There was no in-hospital mortality in EVAR group. In the OAR group, 2/146 (1.4%) patients
255
died postoperatively during the index admission period (P = 0.521). Of these, one patient
256
suffered from severe pancreatitis and died of multisystem organ failure subsequent to septic
257
shock on postoperative day 76. The other patient with symptomatic AAA of 6.7cm in
258
diameter received an urgent operation. Subsequently, graft thrombosis with dissection at the
259
proximal clamping site occurred and this was detected during a reoperation. Despite
260
undergoing thrombectomy, the patient could not be recovered from reperfusion injury and
261
died 1 day after the reoperation.
262
OS rates at 1, 3, and 5 years, respectively, were 94, 78, and 71% in the OAR group, 90, 86,
263
and 77% in the IFU EVAR group, and 93, 56, and 45% in the non-IFU EVAR group (P =
264
0.098, Fig.2). Post-hoc comparisons revealed that the OS rate in the non-IFU EVAR group
265
was significantly lower than in the OAR group (P = 0.031). However, there were no
266
significant differences between the OAR and IFU EVAR group (P = 0.890), or between the
267
IFU EVAR and non-IFU EVAR groups (P = 0.207).
268
After the procedure, in OAR group 13 patients were lost to follow up and 39 patients expired,
269
of whom causes of death were unknown in 7 patients. Other causes of death included heart
270
disease (n=11; CAD, heart failure), malignancy (n=8), cerebral infarction (n=6), pulmonary
271
disease (n=4; COPD, pneumonia), cardiac tamponade due to ascending aortic rupture (n=1),
272
intestinal obstruction (n=1), and reperfusion injury (n=1). Meanwhile, no one was lost to
273
follow up in both of EVAR groups. Causes of death included malignancy (n=3), suicide (n=1),
274
heart failure (n=1), and unknown causes (n=2) in IFU group, while those included pulmonary
275
disease (n=4; pneumonia, COPD), heart failure (n=3), malignancy (n=2), acute myocardial
AC C
EP
TE D
M AN U
SC
RI PT
251
ACCEPTED MANUSCRIPT 276
infarction (n=1), acute renal failure (n=1), peritonitis (n=1), and unknown causes (n=3) in
277
non-IFU EVAR group. After exclusion of deaths of unknown cause, no aneurysm-related
278
death was identified in EVAR group.
RI PT
279
4. DISCUSSION
281
In contrast to the lower perioperative mortality after EVAR, the long-term results of EVAR
282
found in recent RCTs suggest that EVAR does impair survival and leads to higher
283
reintervention rates. This in turn leads to the convergence of survival rates to a similar extent
284
as those seen after open repair.6,8,12,13 However, there is controversy in regard to
285
reinterventions after OAR and EVAR. As described above, the definition of reintervention
286
has differed according to RCTs, and especially whether laparotomy- or wound-related
287
reinterventions after OAR are included. For example, in contrast to previously described
288
RCTs, Schermerhorn et al. reported a comparison of EVAR with OAR for an AAA in the
289
Medicare population, and demonstrated that late reinterventions related to AAA were more
290
common after EVAR but were balanced by an increase in laparotomy-related reinterventions
291
and hospitalizations after OAR.14,15 In our study, RFS and OS rates in non-IFU EVAR group
292
were significantly lower than that of OAR group. However, there were no differences in RFS
293
or OS rates between the OAR and IFU EVAR groups during a mid-term follow-up. In
294
addition, the patients in EVAR were older and had higher rate of comorbidities such as CAD,
295
CHF, and arrhythmia compared with the OAR group. Therefore, EVAR can be safely used for
296
the treatment of patients with unruptured AAA who are suitable for the IFU, although those
297
patients still need to be under close surveillance for detecting delayed endoleaks.
298
Currently, technological advances and accumulated experience have allowed an increasing
299
number of high-risk patients to receive EVAR. In fact, the number of patients who undergo
300
EVAR is on the increase even if the anatomy of the aneurysm is unsuitable for the IFU.
AC C
EP
TE D
M AN U
SC
280
ACCEPTED MANUSCRIPT However, the application of EVAR outside the IFU can cause a potential rupture of the AAA
302
following aneurysm sac enlargement.16 In our study, more reinterventions were conducted in
303
the non-IFU EVAR group than other groups. Final angiography revealed that there were 6
304
Type Ia endoleaks and 3 Type Ib endoleaks in the non-IFU EVAR group. Although the follow
305
up assessment confirmed resolution of Type I endoleak in 6 patients without any
306
reinterventions, 3 of 9 cases required reintervention afterwards and all involved Type Ia
307
endoleaks. In regard to overall survival, the survival rate in both the OAR and non-IFU
308
EVAR groups remained similar at about 90% up to 2 years after operation; however, it fell
309
sharply in the non-IFU EVAR group thereafter. It appears that this was caused by higher
310
proportions of elderly patients and comorbidities. In other words, the OS rate was
311
significantly lower in the non-IFU EVAR group compared with the OAR group after 2 years
312
had passed, and the reintervention rate was higher than in the OAR group during the follow-
313
up period. Therefore, patients treated with non-IFU EVAR should be under strict surveillance
314
for the development of complications.
315
Proximal neck features remain the most important factors determining patient suitability for
316
the IFU for EVAR.9 Although the effect of a short neck on the outcome is obvious, the role of
317
neck angulation is still disputable. As demonstrated in previous studies comparing hostile
318
neck anatomy with a more favorable one, a more angled aortic neck can induce more
319
complications of Type I endoleaks, migrations, adjunctive interventions, and sac
320
expansions.17-20 However, when studying the application of EVAR using the Endurant device
321
with about a 4-year follow-up, primary clinical success and freedom from neck-related
322
complications and reinterventions did not differ according to the extent of angulation.21 While
323
the neck lengths of patients in the above studies also differed between groups in addition to
324
angulation, our patients had similar neck lengths between groups but with more severe
325
angulation in the non-IFU EVAR group. This difference in anatomy appears to have
AC C
EP
TE D
M AN U
SC
RI PT
301
ACCEPTED MANUSCRIPT reinforced the effect of neck angulation on EVAR outcome. In this study, reinterventions
327
were performed in 3 cases treated with IFU EVAR and 11 cases with non-IFU EVAR. Among
328
them, there were no reinterventions caused by proximal attachment problem in the IFU
329
EVAR group, however, 4 (36%) cases of reintervention were caused by Type Ia endoleak in
330
the non-IFU EVAR group. In addition, irrespective of the reasons for reintervention, 5 of 6
331
patients undergoing graft-related reintervention in non-IFU EVAR group had angulated necks.
332
This suggests that neck angulation is an important obstacle to be overcome for better
333
outcomes after non-IFU EVAR.
334
Unlike western countries with frequent short aortic neck in non-IFU EVAR, our study shows
335
that aortic neck angulation acts as an important limitation regarding the IFU of endograft. It is
336
noteworthy that the proportion of female gender was significantly higher in non-IFU EVAR
337
group than other groups (non-IFU EVAR, 25%; OAR, 14%; IFU EVAR, 7%, P=0.043). It
338
turned out that 10 of 13 female patients in non-IFU EVAR group had angulated neck. This
339
finding is similar to the result of previous studies that female gender is less likely to meet the
340
IFU for EVAR owing to unfavorable neck anatomy or poor iliac access vessels.22-24 In
341
addition, more research in order to verify the anatomical variation by race is required. It
342
seems that the effect of race on the outcomes in EVAR has not been much investigated.
343
Existing RCTs usually have been conducted in the west and the study centers were almost
344
located in Europe or United states. Besides, in most RCTs, analysis according to race has not
345
been performed. Even in a few studies mentioning the categorization by race, the Caucasian
346
race forms more than 80% of patients.13 If the outcomes were compared by various races
347
consisting of an equivalent number of patients, more significant results in the relationship
348
between anatomical characteristics and race could be drawn, which may result in remarkable
349
progress in EVAR following accelerated development of devices compatible with a specific
350
race.
AC C
EP
TE D
M AN U
SC
RI PT
326
ACCEPTED MANUSCRIPT One of the limitations of this study is that relatively few patients were enrolled and followed
352
up during a short period of time, and the different average durations among groups. Moreover,
353
fewer patients were assigned to the EVAR procedure than in the OAR group. This must have
354
been caused by the late introduction of the EVAR procedure compared with OAR. Device
355
selection strongly reflected the surgeon’s preference. The selection of endograft was also
356
heavily weighted toward some specific devices. We have mainly used infrarenal fixation
357
devices (Excluder or C3) because of the well-known lower rate of limb occlusion and
358
recently reported additional merits of repositioning after deployment. However, it is known
359
that the use of a familiar device will enhance the outcome of any procedure. Thus, following
360
the surgeon’s preference was not problematic even though it might have influenced the
361
results. Even though the major factors creating a conflict in regard to the execution of EVAR
362
outside the IFU are long-term durability and safety, unfortunately, this study did not show
363
that how many patients among 12 demises of unknown cause exactly died of rupture- or
364
aneurysm-related complications. An underestimation of aneurysm related complication or
365
mortality may have occurred because of the lack of information about cause of death.
366
Whether performing EVAR in patients outside the IFU is ethical is a controvertible issue.
367
However, it is well known that there are no definite alternatives for patients who have
368
prohibitive risk for operation and concurrently do not satisfy the IFU of commercial EVAR
369
devices.25 Even though patients did not meet the IFU, we often decided to perform EVAR
370
instead of OAR on account of their poor general conditions. However, as in our series, non-
371
IFU EVAR has been known to be associated with higher rate of failed EVAR and
372
complications during follow-up. Therefore, risk of open surgery, risk and long-term
373
complications after EVAR in unsuitable anatomy, patients’ preferences, and life expectancy
374
should be considered and discussed thoroughly. This is not a randomized, prospective study,
375
but a retrospective design in a single center, so the results should be interpreted carefully and
AC C
EP
TE D
M AN U
SC
RI PT
351
ACCEPTED MANUSCRIPT 376
without making generalizations. Nonetheless, there is importance in mirroring real-world
377
surgical practice without manipulating it artificially.
378
5. CONCLUSIONS
380
In the IFU EVAR group, the RFS and OS rates were similar to the OAR group during mid-
381
term follow-up. Meanwhile, in the non-IFU group, the RFS and OS rates were lower than in
382
the OAR group and final endoleak was more common than in the IFU EVAR group.
383
Although not statistically significant, reintervention was more commonly performed in the
384
non-IFU EVAR than in the IFU EVAR group. Therefore, performing EVAR in non-IFU
385
situations should be planned carefully and under strict surveillance for the development of
386
complications.
387
TE D EP
389
6. ACKNOWLEDGEMENT
AC C
388
M AN U
SC
RI PT
379
ACCEPTED MANUSCRIPT 390
REFERENCES
391
1.
abdominal aortic aneurysms. Ann Vasc Surg 1991;5:491-9. 2.
nonagenarians: a systematic review. Ann Vasc Surg 2015;29:385-91.
394 395
Wigley J, Shantikumar S, Hameed W, et al. Endovascular aneurysm repair in
RI PT
392 393
Parodi JC, Palmaz JC, Barone HD. Transfemoral intraluminal graft implantation for
3.
Dangas G, O'Connor D, Firwana B, et al. Open versus endovascular stent graft repair of abdominal aortic aneurysms: a meta-analysis of randomized trials. JACC
397
Cardiovasc Interv 2012;5:1071-80. 4.
elective abdominal aortic aneurysm repair. J Vasc Surg 2013;57:1676-83, 83 e1.
399 400
Qadura M, Pervaiz F, Harlock JA, et al. Mortality and reintervention following
M AN U
398
SC
396
5.
Becquemin JP. The ACE trial: a randomized comparison of open versus endovascular repair in good risk patients with abdominal aortic aneurysm. J Vasc Surg
402
2009;50:222-4; discussion 4. 6.
repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1881-9.
404 405
De Bruin JL, Baas AF, Buth J, et al. Long-term outcome of open or endovascular
7.
Lederle FA, Freischlag JA, Kyriakides TC, et al. Outcomes following endovascular vs
EP
403
TE D
401
open repair of abdominal aortic aneurysm: a randomized trial. JAMA 2009;302:1535-
407
42.
408
8.
9.
Lee JT, Ullery BW, Zarins CK, et al. EVAR deployment in anatomically challenging
necks outside the IFU. Eur J Vasc Endovasc Surg 2013;46:65-73.
411 412
United Kingdom ETI, Greenhalgh RM, Brown LC, et al. Endovascular versus open
repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1863-71.
409 410
AC C
406
10.
Walker J, Tucker LY, Goodney P, et al. Adherence to endovascular aortic aneurysm
413
repair device instructions for use guidelines has no impact on outcomes. J Vasc Surg
414
2015;61:1151-9.
ACCEPTED MANUSCRIPT 415
11.
Zandvoort HJ, Goncalves FB, Verhagen HJ, et al. Results of endovascular repair of
416
infrarenal aortic aneurysms using the Endurant stent graft. J Vasc Surg 2014;59:1195-
417
202. 12.
Becquemin JP, Pillet JC, Lescalie F, et al. A randomized controlled trial of
RI PT
418 419
endovascular aneurysm repair versus open surgery for abdominal aortic aneurysms in
420
low- to moderate-risk patients. J Vasc Surg 2011;53:1167-73 e1. 13.
Lederle FA, Freischlag JA, Kyriakides TC, et al. Long-term comparison of
SC
421
endovascular and open repair of abdominal aortic aneurysm. N Engl J Med
423
2012;367:1988-97.
424
14.
M AN U
422
Schermerhorn ML, O'Malley AJ, Jhaveri A, et al. Endovascular vs. open repair of
425
abdominal aortic aneurysms in the Medicare population. N Engl J Med 2008;358:464-
426
74.
Aortic Aneurysm in the Medicare Population. N Engl J Med 2015;373:328-38.
428
16.
Endovascular Abdominal Aortic Aneurysm Repair. Ann Vasc Surg 2016;31:229-38.
430 431
17.
neck anatomy. J Vasc Surg 2013;57:527-38.
433
18.
Hobo R, Kievit J, Leurs LJ, et al. Influence of severe infrarenal aortic neck angulation on complications at the proximal neck following endovascular AAA repair: a
435
EUROSTAR study. J Endovasc Ther 2007;14:1-11.
436 437
Antoniou GA, Georgiadis GS, Antoniou SA, et al. A meta-analysis of outcomes of endovascular abdominal aortic aneurysm repair in patients with hostile and friendly
432
434
Dingemans SA, Jonker FH, Moll FL, et al. Aneurysm Sac Enlargement after
EP
429
Schermerhorn ML, Buck DB, O'Malley AJ, et al. Long-Term Outcomes of Abdominal
TE D
15.
AC C
427
19.
Malas MB, Jordan WD, Cooper MA, et al. Performance of the Aorfix endograft in
438
severely angulated proximal necks in the PYTHAGORAS United States clinical trial.
439
J Vasc Surg 2015;62:1108-17.
ACCEPTED MANUSCRIPT 440
20.
patients with hostile neck anatomy. Eur J Vasc Endovasc Surg 2012;44:556-61.
441 442
Stather PW, Sayers RD, Cheah A, et al. Outcomes of endovascular aneurysm repair in
21.
Oliveira NF, Bastos Goncalves FM, de Vries JP, et al. Mid-Term Results of EVAR in Severe Proximal Aneurysm Neck Angulation. Eur J Vasc Endovasc Surg 2015;49:19-
444
27.
445
22.
RI PT
443
Sweet MP, Fillinger MF, Morrison TM, et al. The influence of gender and aortic aneurysm size on eligibility for endovascular abdominal aortic aneurysm repair. J
447
Vasc Surg 2011;54:931-7. 23.
Dubois L, Novick TV, Harris JR, et al. Outcomes after endovascular abdominal aortic
M AN U
448
SC
446
449
aneurysm repair are equivalent between genders despite anatomic differences in
450
women. J Vasc Surg 2013;57:382-9 e1.
451
24.
Hultgren R, Vishnevskaya L, Wahlgren CM. Women with abdominal aortic aneurysms have more extensive aortic neck pathology. Ann Vasc Surg 2013;27:547-
453
52.
456 457
Rutherford RB. Management of abdominal aortic aneurysms: which risk factors play a role in decision-making? Semin Vasc Surg 2008;21:124-31.
EP
455
25.
AC C
454
TE D
452
ACCEPTED MANUSCRIPT 458
Table I. Characteristics of patients with open aneurism repair (OAR) and endovascular
459
aneurysm repair (EVAR) IFU EVAR
Non-IFU
(n = 146)
(n = 42)
EVAR
P
RI PT
OAR
(n = 52) 69.4 ± 7.8
72.2 ± 7.3
73.2 ± 6.7
0.003
Female
20 (14%)
3 (7%)
13(25%)
0.043
HTN
96 (66%)
24 (57%)
DM
16 (11%)
CAD
31 (21%)
CHF
7 (5%)
Arrhythmia
11 (8%)
CVD
18 (12%)
COPD Renal
0.889
13 (31%)
25 (48%)
0.001
5 (12%)
7 (14%)
0.058
12 (29%)
10 (20%)
0.001
2 (5%)
9 (17%)
0.177
22 (15%)
8 (19%)
13 (25%)
0.270
13 (9%)
4 (10%)
10 (19%)
0.120
2 (5%)
2 (4%)
1.000
TE D
7 (14%)
6 (4%)
AC C
CLD
0.453
5 (12%)
EP
insufficiencya
36 (69%)
M AN U
Comorbidities:
SC
Age, years
460
IFU, instructions for use; HTN, hypertension; DM, diabetes mellitus; CAD, coronary artery
461
disease; CHF, congestive heart failure; CVD, cerebrovascular disease; COPD, chronic
462
obstructive pulmonary disease; CLD, chronic liver disease.
463
464 465
a
Renal insufficiency was defined as preoperative serum creatinine level of 1.5 mg/dL or
higher.
ACCEPTED MANUSCRIPT 466
Table II. Aneurysm characteristics in endovascular aneurysm repair (EVAR) groups treated
467
according to the instructions for use (IFU EVAR), not according to the IFU (non-IFU EVAR),
468
and anatomic causes of unsuitability for the IFU in the non-IFU EVAR group. Non-IFU EVAR
(n = 42)
(n = 52)
Characteristics of aneurysms 14.9 (8.8-31.6)
39.1 (21.3-53.0)
0.000
Infrarenal angulation (°) (IQR)
31.2 (26.3-42.7)
67.0 (35.9-77.4)
0.000
Neck length (mm) (±SD)
32.6 (±10.5)
31.1 (±10.9)
0.504
Neck diameter (mm) (±SD)
21.1 (±2.4)
20.7 (±2.4)
0.457
AAA maximum diameter (mm) (±SD)
55.5 (±8.4)
58.4 (±7.2)
0.079
Distal aortic diameter (mm) (±SD)
25.6 (±7.1)
29.5 (±9.3)
0.031
13.9 (12.2-15.7)
14.7 (11.0-18.4)
0.525
14.2 (12.3-15.9)
14.2 (11.7-16.7)
0.981
CIA length, R. (mm) (±SD)
42.6 (±17.7)
34.9 (±16.4)
0.031
CIA length, L. (mm) (±SD)
45.0 (±19.5)
39.9 (±17.3)
0.184
Iliac tortuosity index, right side (±SD)
1.36 (±0.13)
1.35 (±0.12)
0.549
Iliac tortuosity index, left side (±SD)
1.34 (±0.14)
1.35 (±0.13)
0.713
AC C
EP
TE D
CIA diameter, L. (mm) (IQR)
M AN U
CIA diameter, R. (mm) (IQR)
SC
Suprarenal angulation (°) (IQR)
P
RI PT
IFU EVAR
Anatomic causes of IFU unsuitability Neck causea
35 (67%)
Angulated neck
29 (82%)
Iliac causeb
30 (57%)
Short CIA
14 (46%)
Aneurysm of CIA or EIA
10 (33%)
Infrarenal fixation device
34 (81%)
36 (69%)
ACCEPTED MANUSCRIPT Suprarenal fixation device
8 (19%)
16 (31%)
469
IQR, interquartile range; SD, standard deviation; AAA, abdominal aortic aneurysm; CIA,
470
common iliac artery; EIA, external iliac artery.
473 474
“Neck causes” in this row comprised a short neck (n = 4) and conical neck (n = 2), not
including an angulated neck. b
RI PT
472
a
“Iliac causes” in this row comprised a small EIA (n = 3), occlusion of the CIA or EIA (n =
2), and a narrow aortic bifurcation (n = 1).
SC
471
AC C
EP
TE D
M AN U
475
ACCEPTED MANUSCRIPT 476
Table III. Findings of completion angiography and subsequent treatments in endovascular
477
aneurysm repair groups treated according to the instructions for use (IFU EVAR), or not
478
according to the IFU (non-IFU EVAR).
(n = 42)
(n = 52)
5 (12%)
12 (23%)
Open conversion
1
Palmaz stent
1 2
5
Balloon molding
3
5
Type Ib endoleak
3 (7%)
Additional extender or cuff 2 Balloon molding
3 (7%)
Balloon molding Kissing stent graft
AC C
Type Ia Type Ib 479 480 481 482 483 484
2
0.162
11 (21%)
0.058
6
5
2 (4%)
0.653
2
1
0 (0%)
EP
Final Type I endoleak
TE D
Type III endoleak
1
M AN U
Aortic cuff
P
RI PT
Non-IFU EVAR
SC
Type Ia endoleak
IFU EVAR
9 (17%) 6 3
0.004
ACCEPTED MANUSCRIPT
FIGURE LEGENDS
486
Fig. 1. Kaplan–Meier plots of reintervention-free survival rates in the OAR, IFU EVAR, and
487
non-IFU EVAR groups.
488
Fig. 2. Kaplan–Meier plots of overall survival rates in the OAR, IFU EVAR, and non-IFU
489
EVAR groups.
RI PT
485
AC C
EP
TE D
M AN U
SC
490
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT