Eur J Vasc Endovasc Surg 15, 406-411 (1998)
Importance of Graft Configuration in Outcome of Endoluminal Aortic
Aneurysm Repair: a 5-year Analysis by The Life Table Method
J. May% G. H. White, W. Yu, R. Waugh, M. S. Stephen, M. Arulchelvam and J. P. Harris
Department of Surgery, University of Sydney and Departments of Vascular Surgery and Interventional Radiology,
Royal Prince Alfred Hospital N H M R C Clinical Trials Centre, University of Sydney, Australia
Aim: The aim of this study was to determine the influence of graft configuration on the outcome of endoluminaI repair
of abdominal aortic aneurysm (AAA).
Methods: The 5-year study period extended from May 1992 to May 1997 and included analysis of patients undergoing
endoluminaI AAA repair in the Jirst 4.5-year period with a minimum follow-up period of 6 months. Between May 1992
and November 1996 136 patients underwent endotuminaI AAA repair. Two patients who had endoluminaI repair of
anastomotic AAA and six patients who had secondary endoluminal repair of AAA were excluded, leaving 128 patients
in the study group, There were 117 males and 11 females with a mean age for the group of 71 years. The configuration
of the grafts was tubular aortic (T) (n =50), tapered aortoiliac/femoraI (AI) (n =24) and bifurcated (B) (n =54). Patient
characteristics and co-morbidities were similar in the three groups. The procedures were performed in the operating room
under radiographic control. Follow-up was complete and consisted of regular physical examination and contrast enhanced
computed tomography. Outcome measures were perioperative mortality rate, need for conversion to open repair, presence
of early and late endoleaks, successful exclusion of AAA from the circulation, and survival. Data were analysed by the
life table method.
Results: There was no significant difference in perioperative mortality for T (4%), AI (4%) and B (5.5%) confi'guration
of endograft. Outcome for T, AL and B configurations was respectively: primary conversion (%) 8, I2, 13; early endoleaks
(n=) 5, 0, 1; late endoleaks (n=) 7, O, 1. The overall incidence of failed procedures througho~tt the study period was
higher in tube grafts compared with non-tube (aortoiliac and bifurcated) grafts (p<O.05). Kaplan-Meier curves demonstrated
a success probability at 40 months of 50% for tube grafts and 80% for non-tube grafts. However, a comparison of the
time to procedure failure between tube versus non-tube after adjusting for competing risks (death without prior graft
failure) was non-significant (p = 0.14).
Conclusions: The poor mid-term outcome for tube prostheses requires a reassessment of the criteria for selecting this
configuration. It would be unwise to abandon the use of tube prostheses entirely in endoluminaI repair. With increasing
information on mid and long-term outcome of endoluminal AAA repair it is likely that there will be an increasing
acceptance of treating smaller AAA while they are still suitable for treatment by the endoluminal method and most likely
with tube grafts. A tightening of the criteria for using tube prostheses would seem sensible. In particular, the mininzum
length of distal neck required for endoluminal tube graft repair should be increased to the 2-2.5 cm range.
Introduction
Despite increasing use of the endoluminal method in
the treatment of AAA little attention has been paid to
the influence of graft configuration on outcome. In
1994 we reported a better outcome for type I AAA
compared with type II AAA. 1 The morphology of type
I AAA was suitable for repair using a tubular aortic
endograft while that of type II AAA usually required
either tapered aortoiliac or bifurcated configurations.
Type I aneurysms had a proximal neck of 2 cm or
greater in length, a distal neck of 1.5 cm or greater in
*Please address all correspondence to: J. May, Department of Surgery, University of Sydney DOGNew South Wales 2006, Australia.
length and non-tortuohs iliac arteries. Type II AAA
were so classified if they lacked one or more of the
criteria for type I aneurysms.
At the time of the previously mentioned report the
prostheses were all of the first generation, one piece,
unsupported type requiring an introducing sheath of
24-French internal
diameter.
Subsequent
developments in technology have resulted in the availability of second generation m o d u l a r supported
prostheses with smaller and more flexible introducing
systems. These have enabled the more complex type II
AAA to be treated successfully with greater regularity
than previously. At the same time it was noted that
some of the earlier tube aortic grafts were beginning
to undergo procedure failure. It appeared, therefore,
1078-5884/98/050406 ÷06 $12.00/0 © 1998 W.B.Saunders Company Ltd.
Endoluminal AAA Repair: Outcome by Graft Configuration
407
Table 1. Patient characteristics by group.
Patient characteristics
Tube
aortic
Tapered
AI
Bifurcated
Mean age (years)
Sex (M/F)
Diameter AAA (meanin cm)
Ischaemicheart disease (%)
Hypertension (%)
Diabetes mellitis (%)
Renal impairment(%)
70
47/3
5.3
32 (64)
17 (34)
4 (8)
3 (6)
71
24/0
5.4
16 (67)
8 (33)
3 (12)
3 (12)
70
46/8
5.3
33 (61)
18 (33)
5 (9)
4 (7)
to be an appropriate time to review the influence of
configuration in the outcome of endoluminal AAA
repair. We present our experience with endoluminal
repair of AAA using three different graft configurations with analysis by the life table method over
a 5-year period.
Methods
The study period extended from May 1992 to May
1997 and included analysis of patients undergoing
endoluminal repair in the first 4.5-year period with
follow-up extending from a minimum of 6 months 5 years. Between May 1992 and November 1996, 136
patients underwent endoluminal AAA repair at the
Royal Prince Alfred Hospital. Two patients who had
endoluminal repair of anastomotic AAA and six
patients who had secondary endoluminal repair of
AAA were excluded, leaving 128 patients in the study
group. There were 117 males and 11 females with a
mean age of 71 years. The configuration of the grafts
was tubular aortic (T) (n=50), tapered aortoiliac/
femoral (AI) (n =24) and bifurcated (B) (n = 54). The
patient characteristics for the three groups are listed
in detail in Table 1.
Co-morbidities sufficiently severe to exclude the
patient from open AAA repair at major vascular
centres were present in 53 patients. The nature of these
co-morbidities and their distribution within the three
groups is listed in detail in Table 2. Detailed imaging
by contrast-enhanced computed tomography (CT) and
aortography were performed preoperatively in all
patients. Anatomic criteria for endoluminal grafting
based on these investigations included a proximal
aneurysm neck of length 15 mm or greater and diameter of 28 mm or less and no evidence of a patent
inferior mesenteric artery arising from the aneurysms.
In addition a distal neck of 1.5 cm length between the
AAA and the aortic bifurcation was required when
the tubular aortic configuration was used.
Technique of endoluminal AAA repair
The programme of endoluminal grafting has been
approved by the Institutional Research Review Board
and Ethics Committee. Informed consent was obtained
from each patient. All procedures were performed in
the operating room with patients draped for conventional open repair in the event of failed endoluminal repair. The endografts used were a
modification of the Parodi device (n = 10), the EVT
endograft (EndoVascular Technologies, Menlo Park,
California) (n = 15), the White-Yu endograft (n =78),
the Chuter device (n = 1), the Bard device (n = 1), the
Stentor (Mintec)/Vanguard (Boston Scientific) (n = 23).
The aortoiliac/femoral configuration was combined
with interruption of the contralateral common iliac
artery using either a detachable balloon, blind ended
covered stent or surgical ligation. These manoeuvres
excluded the aneurysm sac from retrograde collateral
filling. A femorofemoral crossover graft was required
to revascularise the contralateral lower limb. Prophylactic antibiotics were administered at the time of
induction of anaesthesia. All patients were anticoagulated with 5000 units of heparin administered
intravenously after exposure of the access artery. Access was via the femoral arteries in 108 patients and
the common iliac arteries via an extra-peritoneal approach in 20 patients. The techniques for transfemoral
and iliac approaches have been described in detail
previously. 2-9 Fluoroscopic monitoring was used for
the delivery and deployment of endografts. Completion angiography was undertaken in all patients.
Follow-up
Follow-up was completed over a 5-year period ending
in May 1997. This consisted of physical examination
and contrast-enhanced CT within the first 10 days
after operation, at 6 months, 12 months and 18 months
after operation, and then annually thereafter. Outcome
Eur J VascEndovascSurg Vol 15, May 1998
J. May et al.
408
Table 2. Co-morbidities.
Tube
aortic
Severe cardiac disease
ASA stage 3b or 4
Renal failure
On dialysis
Successfullytransplanted
Severe respiratory disease
Boushy category III
Chronic liver disease/
portal hypertension
Hostile abdomen
Total
Proportion and % in group
with co-morbidities
Table
Tapered
AI
11
5
1
1
Bifurcated
11
1
1
2
3
2
5
2
3
5
18
13
22
18/50(36)
13/24(54)
22/54(41)
3. Outcome for the three configurations of endograft.
Patients
Primary conversion (%)
Secondary conversion (%)
Endoleaks - early
late
Perioperative mortality (%)
Tube
aortic
Tapered
AI
Bifurcated
50
4 (8)
7 (15)
5
7
2 (4)
24
3 (12)
0 (0)
0
0
1 (4)
54
7 (13)
2 (4)
1
1
3 (5.5)
measures were perioperative mortality rate, need for
conversion to open repair, presence of early and late
endoleaks, 1° successful exclusion of AAA from the
circulation and survival. Successful endoluminal repair was defined as continuing graft function without
endoleak or conversion to open repair and the absence
of an increase in m a x i m u m transverse diameter of
AAA. Persistent endoleak was classified as a procedure
failure, irrespective of whether successful secondary
endoluminal repair was subsequently achieved or not.
Results
Mortality
Deaths occurred in the perioperative period (30 days)
in two patients (4%) following tube aortic graft repair,
one patient (4%) following tapered aortoiliac repair
and three patients (5.5%) following bifurcated graft
repair.
Outcome of operation
Successful endoluminal repair was achieved in 46 of
50 (92%) tube aortic grafts, 21 of 24 (88%) tapered
Eur J Vasc Endovasc Surg Vol 15, May 1998
aortoiliac grafts and 47 of 54 (87%) bifurcated grafts.
Conversion to open repair was undertaken in the
unsuccessful grafts in all configurations. The causes
of failure leading to primary conversion have been
reported in detail previously. 11 Secondary conversion
to open repair on a subsequent occasion was required
in nine patients. Seven of these occurred in tube aortic
grafts. The outcome for the three configurations of
endograft are summarised in Table 3.
Endoleak
Early (within 10 days of operation) endoleaks occurred
in five patients (11%) with tubular aortic grafts, no
patients (0%) with tapered aortoiliac grafts, and one
patient (2%) with a bifurcated graft. Late endoleaks
occurred in seven patients (15%) following tube aortic
graft repair, no patients (0%) with tapered aortoiliac
repair and one patient (2%) with bifurcated repair. The
14 patients with endoleaks were treated by secondary
conversion to open repair in six patients and secondary
endoluminal repair in four patients. Two of the remaining four are being treated conservatively in the
hope that a spontaneous seal will be achieved, and
the other two developed unrelated medical problems
(stroke, hepatic malignancy) of such severity that no
Endoluminal AAA Repair: Outcome by Graft Configuration
Table 4. Procedure failure according to configuration of prostheses.
Tube aortic
Success
Failure
Total
32
18
50
Non-tube
(tapered AI
bifurcated)
Total
65
13
78
97
31
128
409
Graft success (after adjusting for death without prior graft failure)
= 1.0
Aortoiliac
0.8
Bifurcated
0~
~ 0.6
O
-~ 0,4
Chi-squared 5.19, p<0.05.
e 0.2
Numbers at risk
24
16
54
12
50
0
further intervention was undertaken. There were no
deaths in patients undergoing secondary conversion.
27
10
20
30
Months since operation
1
4
'
40
50
Fig. 2. Kaplan-Meier curves for time to procedure failure for three
configurations of grafts after adjusting for death without prior
procedure failure.
Table 5. Treatment and causes of procedure failure in 18 tube
aortic prostheses.
Failed procedures
The results confirmed the clinical impression that the
largest number of failed procedures occurred in the
tube aortic graft group. The overall incidence of failed
procedures throughout the study period was significantly higher in the tube aortic grafts compared
with the non-tube (tapered aortoiliac and bifurcated)
grafts (Table 4). The Kaplan-Meier curves for time to
procedure failure for tube and non-tube grafts are
shown in Fig. 1. A comparison of the time to procedure
failure between tube versus non-tube grafts after adjusting for competing risks 12 (death without prior procedure failure) was non-significant (p = 0.14).
Kaplan-Meier curves for the three configurations of
graft after adjusting for death without prior procedure
Primary conversion to open repair
All technical problems
Secondary conversion to open repair
Short distal neck
Diameter size mismatch
Renal ostium covered
Increasing AAA diameter
Secondary endoluminal repair
Short distal neck
Component separation
Length size mismatch
Untreated endoleaks
Short distal neck
Graft attachment failure
Length size mismatch
4
1
1
1
1
2
1
1
1
1
Table 6. Treatment and causes of procedure failure in three tapered
aortoiliac prostheses.
Graft success (after adjusting for death without prior graft failure)
1.0
Primary conversion to open repair all
technical problems
Secondary conversion to open repair
Secondary endoluminal repair
Untreated endoleaks
Aortoiliac/bifurcated
0.8
~
'
0.6 -
t..-
7
I
0.4
~
m
Numbers at risk
28
0.2 - 78
50
27
I
0
10
L
±ube
4
4
I
20
30
Months since operation
I
40
50
Fig. 1. Kaplan-Meier curves for time to procedure failure for tube
versus non-tube grafts after adjusting for death without prior procedure failure.
failure are shown in Fig. 2. The treatment and causes
of procedure failure in patients with three configurations of prostheses are shown in Tables 5-7.
Of the eighteen patients with tube grafts who had
procedural failures, 13 had their operation in the first
2-year segment of the study period. Five of these 13
failures were late, occurring in the subsequent 3-year
segment of the study period. Procedural failure was
not related to device type.
Eur J Vasc Endovasc Surg Vol 15, May 1998
J. May et al.
410
Table 7. Treatment and causes of procedure
prostheses.
failure in 10 bifurcated
Primary conversion to open repair
All technical problems
Secondary conversion to open repair
Renal ostium covered
Rupture
Secondary endoluminal repair
Untreated endoleaks
Incomplete seal,
contralateral limb
Discussion
The statistical analysis of the data requires further
comment. Although the proportion of tube grafts
failing was significantly greater than non-tube grafts
using the Chi-squared test, analysis of the two groups
by the life table method did not reveal a significant
difference. This was a little surprising considering the
vastly different shape of the Kaplan-Meier curves with
success probabilities at 40 months of 80% for non-tube
grafts and 50% for tube grafts. There are two possible
explanations. There was a disproportionately greater
number of tube grafts performed in the early part of
the study and a disproportionately greater number of
bifurcated grafts in the later part of the study. There
was therefore a longer time period in which the tube
graft procedures could fail compared with the nontube graft procedures. A second explanation is that
the sample size may not be sufficiently large, especially
in the later months of follow-up, to produce a significant difference.
A competing risk approach was used to take into
account the fact that 10 patients died without prior
procedure failure. This gives a truer view of procedure
failure by addressing the fact that some individuals
will never have a procedure failure because they have
been irretrievably removed by another risk, namely
death. Without adjustment, successful outcome would
be falsely inflated by this group of patients.
The poor mid-term outcome for tube aortic prostheses requires analysis of the timing and causes of
failure in these prostheses. The proportion of failures
13/27 (48%) occurring in patients who had their operations in the first 2-year period compared with those
having their operation in the subsequent 3-year period
5/23 (22%) emphasises the importance of the learning
curve in technical skill and device design. The fourtube prostheses which required primary conversion
to open repair resulted from technical problems in the
early phase of the study (Table 5). Of the seven patients
requiring secondary conversion five had persistent
endoleaks. The cause of these persistent endoleaks
was judged to be an inadequate distal neck or cuff in
Eur J Vasc Endovasc Surg Vol 15, May 1998
four. In the fifth patient there was a size mismatch in
the proximal and distal neck due to a prosthesis of
inadequate diameter. The remaining two patients required secondary conversion for inadvertent covering
of the renal ostia in one and progressive increase in
maximum transverse diameter of AAA in the absence
of any demonstrable endoleak in the other. Of the four
patients undergoing secondary endoluminal repair,
the failure of the primary endoluminal repair was
considered to be due to an inadequate distal neck or
cuff in one, a size mismatch resulting from a prosthesis
of inadequate length in one, and separation of component parts of a modular prosthesis in the remaining
two. In the three patients with untreated endoleaks,
the cause of failure of endoluminal repair was con~
sidered to be a distal neck or cuff of inadequate length
in one, a size mismatch resulting from a prosthesis of
inadequate length in one and a graft attachment device
failure in one.
A distal neck of inadequate length was considered
to be responsible for failure in six of 18 patients
and was therefore the commonest cause of failure
for tubular prostheses. These results suggest that a
reassessment of the criteria for selecting the tubular
configuration is required. It would seem that the minim u m length of the distal neck required for endoluminal tube graft repair should be increased to the
2.2 cm range. The other option would be to avoid or
minimise the use of tube prostheses in the management
of AAA when the endoluminal method is contemplated. From published reports 13'I4it would appear
that some groups have taken this course, although we
are unaware of any reports demonstrating a difference
in outcome between tubular and non-tubular grafts.
While this approach is logical, it is probably unwise
to abandon the use of tube prostheses in endoluminal
repair at this stage. With increasing information on
the mid- and long-term outcome of endoluminal AAA
repair it is likely that there will be an increasing
acceptance of treating smaller AAA while they are
still suitable for treatment by the endoluminal method.
These smaller AAA are more likely to be amenable to
treatment with tube prostheses. Raithel ~s already has
an extensive experience with this approach to AAA
management. In his experience of 205 endoluminal
repair procedures, 133 (65%) were performed with
tube prostheses compared with 70 (34%) bifurcated
and two (1%) aortoiliac prostheses. It must also be
remembered that operations involving the use of bifurcated grafts are more complex procedures and are
more liable to a variety of complications. These include
graft limb stenosis, graft limb thrombosis and graft
limb detachment. In addition, radiation exposure and
Endoluminal AAA Repair: Outcome by Graft Configuration
c o n t r a s t l o a d m a y b e greater. S o m e of these c o m p l i c a t i o n s c a n b e e x p e c t e d to r e s u l t i n p r o c e d u r e failure
with longer follow-up, and indeed we have observed
this i n o n e p a t i e n t after the t e r m i n a t i o n of the s t u d y
period.
T h e t e c h n i q u e of t u b e graft r e p a i r u s i n g the e n d o l u m i n a l m e t h o d looks d e c e p t i v e l y s i m p l e , b u t is q u i t e
d e m a n d i n g . Exact m e a s u r e m e n t of the r e n a l a r t e r y to
aortic b i f u r c a t i o n d i s t a n c e is critical to the o u t c o m e .
I n a d d i t i o n , a l l o w a n c e h a s to b e m a d e for a n y tort u o s i t y i n this s e g m e n t . The m a r g i n for e r r o r i n
d e p l o y i n g the p r o s t h e s i s i n a p o s i t i o n m o r e s u p e r i o r
to the b i f u r c a t i o n t h a n p l a n n e d or d e p l o y i n g the p r o s thesis i n the c o m m o n iliac a r t e r y is small. It w o u l d
s e e m p r u d e n t , therefore, for v a s c u l a r s u r g e o n s to
m a i n t a i n their skills w i t h the t u b u l a r t e c h n i q u e . This
c o u l d b e d o n e w i t h o u t d e t r i m e n t to the p a t i e n t b y
t i g h t e n i n g the criteria for u s i n g the t u b u l a r prosthesis.
References
1 MAY J, WHITEGH, Yu W, WAUGHRC, STEPHENMS, HARRISJP.
Results of endoluminal grafting of abdominal aortic aneurysms
are dependent on aneurysm morphology. Ann Vasc Surg 1996;
10: 254=226. (Presented at the Joint Meeting of the Society for
Vascular Surgery and the North American Chapter of the Society
for Cardiovascular Surgery, Seattle, 1994).
2 PARODIJC, PALMAZlC, BARONEHD. Transfemoral intraluminal
graft implantation for abdominal aortic aneurysms. Ann Vasc
Surg 1991; 5: 491-499.
3 MAY J, WHITE GH, Yu W, WAUGI~RC, HARRISJP. Treatment
of complex abdominal aortic aneurysms by a combination of
endoluminal and extraluminal aortobifemoral grafts. Ann Vasc
Surg 1994; 19: 924-933.
4 MOOREWS. Endovascular grafting technique: a feasibility study.
411
In: YAO JST, PEARCE WH eds. Aneurysms: New Findings and
Treatments. Norwalk, Conn: Appleton & Lange, 1994; 333-340.
5 YusuP SW, BAKERDM, CHUTERTAM, WHITAKERSC, WENHAM
PR, HOPKINSON BR. Transfemoral endoluminal repair of abdominal aortic aneurysm with bifurcated graft. Lancet 1994; 344:
650-651.
6 SCOTT RA, CHUTER TA. Clinical endovascular placement of
bifurcated graft in abdominal aortic anenrysm without laparotomy. Lancet 1994; 343: 413.
7 WHITE GH, Yu W, MAY J, STEPHENMS, WAUGHRC. A new
non-stented balloon expandable graft for straight or bifurcated
endoluminal bypass. J Endovasc Surg 1994; 1: 16-24.
8 CHUTERTAM, WENDTG, ]7{OPKINSONBR, SCOTTRAB, REISBERG
B, WALKERPJ, WHITEGH. Transfemoral insertion of a bifurcated
endovascular graft for aortic aneurysm repair: the first 22
patients. J Cardiovascular Surg 1995; 3: 121-128.
9 BALMR, EIKELBOOMBC, MAYJ, BELLPRF, SWEDENBORGJ, COLIN
J. Early experience with Transfemoral Endovascular Aneurysm
Management (TEAM) in the treatment of aortic aneurysms.
European J Vasc Endovasc Surg 1996; 11: 214-220.
10 WHITE GH, MAY J, Yu W. "Endoleak" - a proposed new terminology to describe incomplete aneurysm exclusion by an
endoluminal graft. J Endovasc Surg 1996; 3: 124-125.
11 MAYJ, WHITEGH, Yu W, WAUGHRC, McGAHANT, STEPHENMS,
HARRISJP. Endoluminalgrafting of abdominal aortic aneurysms:
causes of failure and their prevention. J Endovasc Surg 1994; 1:
44-52.
12 PEPE MS. Inference with dependent risks in multiple endpoint
studies. J American Statistical Assoc 1991; 86: 770-778.
13 BLUM V~ VOSHAGEG, LAMMERJ, BEYERSDORFF~ TOLLNERD,
KRETSCHMER G, SPILLNER G, POLTERAUERP, NAGEL G, HOLZENBEINT, THURNERS, LANGERM. Endoluminal stent-grafts for
infrarenal abdominal aortic aneurysms. N Engl [ Med 1997; 336:
13-20.
14 YusuP SW, WHITAKERSC, CHUTERTAM, IVANCEVK, BAKERDM,
GREGSON RHS, TENNANTWG, WENHAMPW, HOPKINSONBR.
Early results of endovascular aortic aneurysm surgery with
aortouniiliac graft, contralateral iliac occlusion, and femorofemoral bypass. J Vasc Surg 1997; 25: 165-172.
15 SCI-IUNN CD, HEILBERGERP, KRAUSSM, RITTERW, RAITHELD.
Aortic aneurysm size and graft behaviour after endovascular
stentgrafting - two years of clinical experience and followup observations. Unpublished work presented at the ISCVS
Meeting, London, Septem.ber 1997.
Accepted 7 November 1997
Eur J Vasc Endovasc Surg Vol 15, May 1998