From the Western Vascular Society
`
`CLINICAL RESEARCH STUDIES
`
`Explant analysis of AneuRx stent grafts:
`Relationship between structural findings and
`clinical outcome
`
`Christopher K. Zarins, MD,a Frank R. Arko, MD,a Tami Crabtree, MS,a Daniel A. Bloch, PhD,a
`Kenneth Ouriel, MD,b Robert C. Allen, MD,c and Rodney A. White, MD,d Stanford, Calif; Cleveland,
`Ohio; Naples, Fla; and Los Angeles, Calif
`
`Objective: We reviewed the structural findings of explanted AneuRx stent grafts used to treat abdominal aortic aneurysms,
`and relate the findings to clinical outcome measures.
`Methods: We reviewed data for all bifurcated AneuRx stent grafts explanted at surgery or autopsy and returned to the
`manufacturer from the US clinical trial and worldwide experience of more than 33,000 implants from 1996 to 2003.
`Devices implanted for more than 1 month with structural analysis are included in this article. Explant results were
`analyzed in relation to cause of explantation and pre-explant evidence of endoleak, enlargement, or device migration.
`Results: One hundred twenty explanted stent grafts, including 37 from the US clinical trial, were analyzed. Mean implant
`duration was 22 ⴞ 13 months (range, 1-61 months). Structural abnormalities included stent fatigue fractures, fabric
`abrasion holes, and suture breaks. The mean number of nitinol stent strut fractures per explanted device was 3 ⴞ 4, which
`represents less than 0.2% of the total number of stent struts in each device. The mean number of fabric holes per explanted
`device was 2 ⴞ 3, with a median hole size of 0.5 mm2. Suture breaks were seen in most explanted devices, but composed
`less than 1.5% of the total number of sutures per device. “For cause” explants (n ⴝ 104) had a 10-month longer implant
`duration (P ⴝ .007) compared with “incidental” explants (n ⴝ 16). “For cause” explants had more fractures (3 ⴞ 5; P ⴝ
`.005) and fabric holes (2 ⴞ 3; P ⴝ .008) per device compared with “incidental” explants, but these differences were not
`significant (P ⴝ .3) when adjusted for duration of device implantation. Among clinical trial explants the number of fabric
`holes in grafts in patients with endoleak (2 ⴞ 3 per device) was no different from those without endoleak (3 ⴞ 4 per
`device; P ⴝ NS). The number of fatigue fractures or fabric holes was no different in grafts in clinical trial patients with
`pre-explant aneurysm enlargement compared with those without enlargement. Pre-explant stent-graft migration was
`associated with a greater number of stent strut fractures (5 ⴞ 7 per device; P ⴝ .04) and fabric holes (3 ⴞ 3 per
`bifurcation; P ⴝ .03) compared with explants without migration. Serial imaging studies revealed inadequate proximal,
`distal, or junctional device fixation as the probable cause of rupture or need for conversion to open surgery in 86% of “for
`cause” explants. Structural device abnormalities were usually remote from fixation sites, and no causal relationship
`between device findings and clinical outcome could be established.
`Conclusions: Nitinol stent fatigue fractures, fabric holes, and suture breaks found in explanted AneuRx stent grafts do not
`appear to be related to clinical outcome measures. Longer term studies are needed to confirm these observations. (J Vasc
`Surg 2004;40:1-11.)
`
`Endovascular repair has gained wide acceptance in the
`treatment of abdominal aortic aneurysm, with favorable
`
`From the Division of Vascular Surgery and Health Research and Policy,
`Stanford University Medical Center,a Stanford, Calif, the Department of
`Vascular Surgery, The Cleveland Clinic Foundation,b Cleveland, The
`Department of Vascular Surgery, The Cleveland Clinic,c Naples, and the
`Division of Vascular Surgery, Harbor-UCLA Medical Center,d Los An-
`geles.
`Competition of interest: The authors are paid consultants for Medtronic and
`may have equity interest in the company.
`Presented at the Eighteenth Annual Meeting of the Western Vascular
`Society, Kohala Coast, HI, Sep 20-23, 2003.
`Reprint requests: Christopher K. Zarins, MD, Stanford University Medical
`Center, Division of Vascular Surgery, 300 Pasteur Dr, H3642, Stanford,
`CA 94305-5642 (e-mail: zarins@stanford.edu).
`0741-5214/$30.00
`Copyright © 2004 by The Society for Vascular Surgery.
`doi:10.1016/j.jvs.2004.03.008
`
`results with a variety of devices.1-6 However, adverse
`events, including aneurysm rupture, migration, endoleak,
`and aneurysm enlargement, are cause for concern.7-11 Ad-
`verse clinical events may result from device structural fail-
`ure, including hook fractures, stent corrosion, stent frac-
`tures, and fabric tears.12-17 However, the relationship
`between structural characteristics of implanted endovascu-
`lar devices, clinical performance, and long-term outcome of
`endovascular aneurysm repair is poorly understood.18 Most
`reports of device mechanical failure have been largely based
`on radiographic findings,19 and few explanted devices have
`been subjected to detailed morphologic and structural
`analysis.20 The purpose of this investigation was to evaluate
`the structural characteristics of explanted AneuRx stent
`grafts and to relate the findings to clinical outcome mea-
`sures.
`
`1
`
`TMT 2121
`Medtronic v. TMT
`IPR2021-01532
`
`

`

`2 Zarins et al
`
`JOURNAL OF VASCULAR SURGERY
`July 2004
`
`material (RPM) fabric was used in 936 patients in the
`clinical trial; in the initial 257 clinical trial patients a less
`densely woven polyester fabric (preRPM fabric) was used
`(Fig 2). Stent rings are joined to each other and to the
`fabric with 5-0 polyester sutures. There are 1619 to 2079
`suture attachments in each bifurcation module.
`Method. All bifurcated AneuRx stent-graft devices
`implanted from June 1996 to July 2003, and subsequently
`explanted and returned to the manufacturer (Medtronic
`Vascular) were reviewed (N ⫽ 161). Devices returned after
`unsuccessful implantation or those in place for less than 30
`days were excluded from this analysis (n ⫽ 28), as were 13
`explanted devices with no structural analysis because of
`histologic sectioning. One hundred twenty explanted de-
`vices from the worldwide experience, including 37 ex-
`planted in the US clinical trial, are included in this report.
`Explant analysis included gross inspection and photog-
`raphy of the “as received” specimen, x-ray imaging, tissue
`digestion and cleaning of the stent graft, magnified optical
`imaging, and ultrastructural analysis of the stent graft.
`Scanning electron microscopy of each stent fracture surface
`was used to differentiate acute overload fractures (induced
`by mechanical trauma during the explant procedure) from
`fatigue fractures (present in vivo before explantation).
`Overload fractures were characterized by micro-ovoid co-
`alescence surface anatomy, whereas
`fatigue fractures
`showed radial lines that originated at the fracture origin and
`fanned out parallel to the direction of crack growth21 (Fig
`2). Graft defects included fabric weave separation, fabric
`abrasion holes, and explantation-induced cuts. The con-
`tour of each fabric defect was traced, and cross-sectional
`area was calculated with Scion Image Software, version
`4.0.2. Fabric holes were defined as defects larger than 0.2
`mm2 in cross-section, the manufacturing specification for
`the graft fabric and the approximate cross-sectional area of
`a 25-gauge needle. Suture abnormalities included fraying
`and broken sutures.
`The primary and secondary reasons for surgical conver-
`sion and stent-graft explantation were determined from the
`explanting surgeon’s operative report and clinical sum-
`mary. The cause of death, and the status of the aneurysm
`and stent graft were determined from the autopsy report.
`All explants retrieved at surgical conversion, as well as those
`retrieved at autopsy where the cause of death was aneurysm
`rupture or stent-graft complication, were grouped as “for
`cause” explants. Explants retrieved at autopsy where the
`cause of death was unrelated to the aneurysm or stent graft
`were considered “incidental” explants. Clinical records,
`and preoperative and post-implantation imaging studies
`were reviewed by two experienced surgeons, independent
`of explant findings, to determine the reason for clinical
`events leading to “for cause” explantation. For the 37
`clinical trial explants, pre-explant imaging evidence of en-
`doleak, aneurysm enlargement, or stent-graft migration, as
`determined by the core laboratory (phase II patients) or
`clinical center (phase III patients), were related to quanti-
`tative explant analysis findings.
`
`Fig 1. Nitinol exoskeleton is composed of individual 1-cm long
`stent rings (top) joined end-to-end and attached to the graft fabric
`with sutures. Each stent diamond has four struts. The early proto-
`type “stiff” bifurcation module (bottom left) had a single 5-cm long
`bifurcation stent. The flexible segmented design (bottom right) was
`used in most clinical trial patients, and is the design used in the
`commercially marketed stent graft.
`
`MATERIAL AND METHODS
`AneuRx stent-graft structure. The AneuRx stent
`graft is composed of an exoskeleton of nitinol stent rings
`sutured to a polyester fabric graft. The bifurcation module
`contains seven aortic rings and 14 to 17 iliac rings, depend-
`ing on the length of the stent graft. There are 728 to 1020
`(mean, 907 ⫾ 82) nitinol diamond strut elements in each
`bifurcation module. The initial bifurcation design con-
`tained a single 5-cm long bifurcated stent segment, which
`resulted in a stiff bifurcation. This was used in the early
`clinical experience and in the first 174 patients in the clinical
`trial. The current segmented bifurcation design, used in
`1019 patients in the clinical trial, includes 1-cm long stent
`rings throughout the stent graft, which results in a flexible
`bifurcation (Fig 1). The woven polyester fabric material
`used in current AneuRx devices was introduced in Decem-
`ber 1997 during the clinical trial. This reduced porosity
`
`

`

`JOURNAL OF VASCULAR SURGERY
`Volume 40, Number 1
`
`Zarins et al 3
`
`Fig 2. A, Explanted device demonstrates a broken suture between aortic stent rings 3 and 4, and a stent strut fracture
`in stent ring 4. B, Scanning electron microscopy image confirms that this stent strut fracture is a fatigue fracture, by the
`fault lines visible on the fracture surface. C, Reduced porosity material (RPM) fabric used in most patients in the clinical
`trial and in the commercially marketed stent graft. D, PreRPM fabric used in the initial devices had a less dense fabric
`weave.
`
`Statistical analysis. Results were expressed as mean ⫾
`SD for continuous data, and as percentage for categorical
`data. Differences among groups were determined with the
`Wilcoxon rank sum test for continuous data, and the ␹2 or
`Fisher exact test for binary comparisons. Analysis of covari-
`ance was performed to adjust for factors that were unbal-
`anced between the groups and could potentially affect the
`outcomes. Differences at P ⬍ .05 were considered statisti-
`cally significant.
`
`RESULTS
`Explant patient population. The explant patient co-
`hort (N ⫽ 120) was similar to the population of patients
`treated in the US clinical trial.1 Mean age in the explant
`patient group at the time of device implantation (72 ⫾ 7
`years) was not different from age in the clinical trial patient
`group (73 ⫾ 8 years). There was no difference in gender
`distribution between the explant (88% men, 12% women)
`and clinical trial groups (89% men, 11% women). Mean
`duration of device implantation before explantation was 22
`⫾ 13 months (range, 1-61 months). The distribution of
`stiff versus flexible bifurcation module stent design in the
`worldwide explant group (5% vs 95%) was significantly
`different from the distribution among clinical trial patients
`(14% vs 86%; P ⫽ .004). The distribution of preRPM versus
`RPM fabric in the worldwide explant group was 11% versus
`89%, significantly different from the distribution in clinical
`trial patients (21% vs 79%; P ⫽ .008).
`
`Reasons for explantation. The primary and second-
`ary reasons for device explantation, as determined by the
`explanting physician, are shown in Table I. Aneurysm
`enlargement with or without endoleak was the most com-
`mon reason for surgical conversion, followed by aneurysm
`rupture and stent-graft migration. Three autopsy explants
`(two in the clinical trial group) were from patients who died
`of aneurysm rupture, and are included among the “for
`cause” rupture explants. There were 104 “for cause” ex-
`plants and 16 “incidental” autopsy explants.
`Nitinol stent fatigue fractures. One or more mitinol
`stent fatigue fractures were observed in 66% of explanted
`devices, and no fractures in 34% of devices. The mean
`number of strut fractures per explanted device was 3 ⫾ 4
`(range, 0-29). Stent strut fractures represented less than
`0.2% of the total number of stent struts in each device. The
`total number of fatigue fractures per explanted device is
`shown in Fig 3. Only six explants (5%) had 10 or more strut
`fractures. The maximum number of strut fatigue fractures
`in a bifurcation module was 13. The maximum number of
`fractures in any module was 29, and were found in an aortic
`extender cuff that was severely angulated and distorted.
`The distribution of stent fractures among modular
`device components is shown in Table II. Most fatigue
`fractures were found in the bifurcation module, with 2 ⫾ 3
`fractures per bifurcation. Bifurcation fractures were associ-
`ated with severe neck angulation, and were usually located
`in aortic stent rings 3 (70% of fractures) and 4 (12% of
`
`

`

`4 Zarins et al
`
`Table I. Reason for graft explanation
`
`Worldwide
`(N ⫽ 120)
`
`Clinical trial
`(N ⫽ 37)
`
`“For cause” explanation
`Rupture*
`Symptomatic AAA
`Component separation or
`migration
`Aneurysm enlargement
`Endoleak or inadequate
`seal
`Unknown
`Symptomatic AAA
`(nonruptured)
`Migration
`Aneurysm enlargement
`Endoleak
`Infection
`Component separation
`Migration
`Aneurysm enlargement
`Endoleak
`Type I
`Type II
`Type III
`Unknown source
`No endoleak
`Endoleak without
`enlargement
`Type I
`Type II
`Type III
`Aortoenteric fistula
`Infection
`Limb occlusion
`“Incidental” autopsy
`explantation
`
`104
`
`17
`3
`3
`
`3
`4
`
`4
`4
`
`1
`1
`1
`1
`4
`14
`39
`31
`6
`14
`6
`5
`8
`11
`
`9
`1
`1
`4
`4
`7
`16
`
`31
`
`7
`0
`0
`
`3
`1
`
`3
`0
`
`5
`2
`15
`12
`4
`2
`3
`3
`3
`0
`
`2
`0
`0
`6
`
`AAA, Abdominal aortic aneurysm.
`*Three grafts in worldwide group and two in clinical trial group were
`explanted at autopsy.
`
`fractures), where the circular aortic stent contour transi-
`tions to a bi-oval configuration. Iliac limb fractures were
`less common, and were localized in areas of severe iliac
`angulation. Fatigue fractures were rarely seen in the seal
`zones of the device to the aortic neck or iliac arteries or in
`the modular iliac junction.
`Fabric holes and weave separation. Fabric weave
`separation was observed in explants with the early preRPM
`fabric, and was most prominent in areas of suture pull and
`fabric stretch in severe angulations and bends. Weave sep-
`aration resulted in spaces between fabric fibers of less than
`0.2 mm2. Fabric abrasion and fabric holes were related to
`wear of the fabric against the apex of metallic stent dia-
`monds in areas of severe angulation and suture breaks.
`Fabric holes were rarely seen in relation to stent strut
`fractures.
`One or more fabric holes larger than 0.2 mm2 were
`observed in 45% of explanted devices, and no fabric holes
`were found in 55% of devices. The mean number of holes
`per explanted device was 2 ⫾ 3 (range, 0-17). Mean fabric
`
`JOURNAL OF VASCULAR SURGERY
`July 2004
`
`hole size was 0.7 ⫾ 0.7 mm2, with a median hole size of 0.5
`mm2 (approximately the size of a 21-gauge needle). The
`total number of fabric holes per explanted device is shown
`in Fig 4. Only four explanted devices (3%) had 10 or more
`holes. The distribution of fabric holes per device module is
`shown in Table II. Fabric holes were most often found in
`the bifurcation module, with 1 ⫾ 2 holes per bifurcation.
`There were no significant differences in the number of stent
`fractures or fabric holes between preRPM fabric explants
`and RPM fabric explants.
`Suture integrity. Broken and abraded sutures were
`found in almost all explanted devices, most commonly in
`areas of severe angulation. Suture breakage was most often
`seen at junction stitches between stent rings (Fig 2), and
`breakage enabled partial stent ring separation or overlap.
`The median number of broken sutures per explanted device
`studied was 22, which was less than 1.5% of the total
`number of sutures in each device.
`Duration of implant. The number of fatigue frac-
`tures and fabric holes as a function of implantation time is
`shown in Fig 5. The number of explants at each time point
`is indicated in the figure. The device with the largest
`numbers of fatigue fractures (29 fractures on one extender
`cuff) had been implanted for 30 months.
`“For cause” versus “incidental” explantation. A
`comparison of “for cause” explantation and “incidental”
`explantation is shown in Table III. Patients in the world-
`wide “for cause” explantation group were 6 years younger
`than patients in the “incidental” explantation group (P ⫽
`.003), and their devices had been implanted 10 months
`longer (P ⫽ .007). Preoperative patient risk factor analysis
`in the 37 clinical trial explants revealed a higher rate of
`symptomatic cardiac arrhythmia (P ⫽ .02) in the “inciden-
`tal” explant group, consistent with documented cardiac
`cause of death in 70% of this group. Explant analysis
`revealed a significantly greater number of stent fatigue
`fractures and fabric holes in “for cause” explants compared
`with “incidental” explants (P ⫽ .005 and P ⫽ .008, respec-
`tively; Wilcoxon rank sum test). However, when analysis of
`covariance was used to adjust for differences in duration of
`implantation, the differences in fatigue fractures and fabric
`holes between the two groups were not significant (P ⫽ .3
`for both).
`Endoleak. Among clinical trial patients endoleak was
`documented on the final imaging study before explantation
`in 26 patients, and was absent in 11 patients. Endoleak rate
`was higher with “for cause” explants (81%) compared with
`“incidental” explants (17%; P ⫽ .005). There was no
`significant difference in endoleak rate between explants
`with fabric holes (76%) compared with those with no fabric
`holes (63%). Similarly, there was no significant difference in
`endoleak rate between explants with stent fractures (78%)
`compared with those with no stent fractures (57%). There
`was no significant difference in the number of fabric holes
`or stent strut fractures between explants from patients with
`endoleak compared with those without endoleak (Table
`IV).
`
`

`

`JOURNAL OF VASCULAR SURGERY
`Volume 40, Number 1
`
`Zarins et al 5
`
`Fig 3. Number of stent strut fatigue fractures per explanted device.
`
`Fig 4. Number of fabric holes (⬎0.2 mm2) per explanted device.
`
`Fig 5. Fractures and holes as a function of number of months the device was implanted. Number of explants at each
`time point is indicated on the x-axis.
`
`Aneurysm enlargement. Among patients in the clin-
`ical trial group, aneurysm enlargement (diameter increase
`⬎5 mm compared with baseline computed tomography
`
`scan) before explantation was documented in 15 patients;
`22 aneurysms exhibited no enlargement. There was no
`significant difference in enlargement rate between explants
`
`

`

`6 Zarins et al
`
`JOURNAL OF VASCULAR SURGERY
`July 2004
`
`Table II. Stent fatigue fractures and fabric holes by device module
`
`Bifurcation
`
`Iliac limb
`
`n
`
`%
`
`n
`
`Worldwide explanations (N ⫽ 120)
`One or more fatigue fractures
`No. of fractures (mean ⫾ SD)
`One or more fabric holes
`No. of holes (mean ⫾ SD)
`Clinical trial explantations (N ⫽ 37)
`One or more fatigue fractures
`No. of fractures (mean ⫾ SD)
`One or more fabric holes
`No. of holes (mean ⫾ SD)
`
`76
`2.4 ⫾ 3.0
`
`43
`1.0 ⫾ 2.0
`
`21
`1.7 ⫾ 2.0
`16
`1.6 ⫾ 2.6
`
`57
`
`43
`
`8
`0.1 ⫾ 0.3
`22
`0.5 ⫾ 1.3
`
`2
`0.1 ⫾ 0.4
`6
`0.5 ⫾ 1.2
`
`%
`
`7
`
`18
`
`5
`
`16
`
`Extender cuff
`
`n
`
`8
`0.5 ⫾ 2.9
`14
`0.4 ⫾ 1.7
`
`4
`0.9 ⫾ 4.8
`2
`0.2 ⫾ 0.9
`
`%
`
`7
`
`12
`
`11
`
`5
`
`Table III. “For cause” versus “incidental” explant analysis
`
`Worldwide explants (N ⫽ 120)
`
`Clinical trial explants (N ⫽ 37)
`
`For cause
`(n ⫽ 104)
`
`71 ⫾ 6.9
`88:12
`
`12:88
`
`24 ⫾ 13
`
`74/104
`71
`3.3 ⫾ 4.5
`
`51/104
`49
`2.2 ⫾ 3.4
`
`No cause
`(n ⫽ 16)
`
`77 ⫾ 7.0
`80:20
`
`6:94
`
`14 ⫾ 12
`
`5/16
`31
`1.1 ⫾ 2.5
`
`3/16
`19
`0.2 ⫾ 0.4
`
`P
`
`.003
`NS
`
`NS
`
`.007
`
`.002
`
`.005*
`
`.02
`
`.008*
`
`For cause
`(n ⫽ 31)
`
`71 ⫾ 7.9
`87:13
`
`23:77
`
`31 ⫾ 9
`
`21/31
`68
`3.2 ⫾ 5.3
`
`20/31
`65
`2.6 ⫾ 3.6
`
`No cause
`(n ⫽ 6)
`
`74 ⫾ 7.5
`67:33
`
`17:83
`
`24 ⫾ 10
`
`2/6
`33
`0.5 ⫾ 0.8
`
`1/6
`17
`0.2 ⫾ 0.4
`
`P
`
`NS
`NS
`
`NS
`
`NS
`
`NS
`
`NS (.06)*
`
`NS
`(.07)
`.03*
`
`Age at implant (y)
`Gender
`(%male:female)
`Pre-RPM:RPM fabric
`(%)
`Implant duration (mo)
`Any device fracture
`n
`%
`Stent fractures per
`device
`Any device hole
`n
`%
`Fabric holes per device
`
`*Based on Wilcoxon rank sum test. When analysis of variance is used to control for duration of implant, P ⫽ .3.
`
`with fabric holes (38%) compared with explants without
`holes (44%). Similarly, there was no significant difference in
`aneurysm enlargement between explants with fatigue frac-
`tures (39%) compared with those with no fractures (43%).
`There was no significant difference in the number of fabric
`holes or strut fractures between explants from patients with
`or without aneurysm enlargement (Table V).
`Device migration. In the clinical trial group, pre-
`explant imaging studies documented stent-graft migration
`in 14 patients; 23 patients had no stent-graft migration.
`Explants with stent fractures in any device module were
`more likely to have migrated (52%) than were explants with
`no fractures (14%; P ⫽ .02). The number of stent fractures
`was higher in explants from patients with stent-graft migra-
`tion than in explants with no migration (P ⫽ .04; Table
`VI). The number of fabric holes was higher in bifurcation
`modules from patients with stent-graft migration (P ⫽
`.03); however, there was no significant difference in the
`total number of fabric holes per device (P ⫽ .08; Table VI).
`Clinical case review. Retrospective review of clinical
`records and sequential imaging studies revealed evidence of
`
`short or absent device fixation length, either proximally at
`the aortic neck, distally to the iliac artery, or at a modular
`junction site in 86% of “for cause” explant cases. This
`review suggested that inadequate device fixation was a
`major cause of clinical events leading to rupture or surgical
`explantation in each case. A case example, demonstrating
`inadequate fixation of the modular iliac limb as the cause of
`aneurysm rupture is shown in Fig 6. No structural abnor-
`malities were found in the left iliac limb, whereas focal
`fatigue fractures and fabric holes were found in the bifur-
`cation module. Most structural abnormalities found in
`explanted devices were remote from stent-graft fixation
`sites, and no causal relationship between device structural
`findings and clinical outcome could be established.
`
`DISCUSSION
`Despite a 7-year clinical experience, little information is
`available on the structural characteristics of implanted An-
`euRx devices, because clinical results are good1,6 and rela-
`tively few devices have been retrieved for analysis. The
`sample of 120 explanted stent grafts in this study is a small
`
`

`

`JOURNAL OF VASCULAR SURGERY
`Volume 40, Number 1
`
`Zarins et al 7
`
`Table IV. Endoleak analysis (clinical trial patients)*
`
`Endoleak (n ⫽ 26)
`
`No endoleak (n ⫽ 11)
`
`Fabric holes per bifurcation module
`Fabric holes per device
`Stent fractures per bifurcation module
`Stent fractures per device
`
`1.4 ⫾ 2.4
`2.1 ⫾ 3.2
`1.9 ⫾ 2.0
`3.3 ⫾ 5.7
`
`1.9 ⫾ 3.1
`2.5 ⫾ 3.9
`1.3 ⫾ 1.9
`1.3 ⫾ 1.9
`
`Values represent mean ⫾ SD.
`*Patients with endoleak present on last imaging study prior to explantation compared with patients without endoleak.
`
`Table V. Enlargement analysis (clinical trial patients)*
`
`Enlargement (n ⫽ 15)
`
`No enlargement (n ⫽ 22)
`
`Fabric holes per bifurcation module
`Fabric holes per device
`Stent fractures per bifurcation module
`Stent fractures per device
`
`1.3 ⫾ 2.2
`1.6 ⫾ 2.8
`1.4 ⫾ 1.8
`3.8 ⫾ 7.3
`
`1.8 ⫾ 2.9
`2.6 ⫾ 3.7
`1.9 ⫾ 2.1
`2.0 ⫾ 2.1
`
`P
`
`NS
`NS
`NS
`NS
`
`P
`
`NS
`NS
`NS
`NS
`
`Values represent mean ⫾ SD.
`*Patients with aneurysm enlargement (⬎5 mm) documented on serial imaging studies before explanation compared with patients with no enlargement.
`
`Table VI. Migration analysis (clinical trial patients)*
`
`Migration (n ⫽ 14)
`
`No migration (n ⫽ 23)
`
`Stent fractures per bifurcation module
`Stent fractures per device
`Fabric holes per bifurcation module
`Fabric holes per device
`
`2.5 ⫾ 2.0
`4.6 ⫾ 7.3
`2.5 ⫾ 3.2
`3.6 ⫾ 4.6
`
`1.2 ⫾ 1.8
`1.6 ⫾ 2.2
`1.0 ⫾ 2.0
`1.4 ⫾ 2.1
`
`Values represent mean ⫾ SD.
`*Patients with evidence of device migration on serial imaging studies before explantation compared with patients without migration.
`
`P
`
`.04
`.04
`.03
`.08
`
`fraction of the 33,000 AneuRx devices implanted world-
`wide. Nonetheless, it is the largest reported systematic
`structural analysis of explanted aortic devices. Although
`there were some early changes in stent design (stiff to
`flexible) and fabric weave (preRPM to RPM), the AneuRx
`device has remained the same over the past 7 years. Thus
`these explant results largely reflect the currently marketed
`device. However, the sample of explants is sharply skewed
`toward the minority of patients with serious adverse events,
`inasmuch as 87% of worldwide explants were retrieved at
`surgical explantation or after aneurysm rupture. Among the
`1193 clinical trial patients, those with aneurysm rupture or
`surgical conversion represented less than 4% of patients,6
`but accounted for 84% of the explants in this study. Device
`recovery rate in clinical trial patients who underwent surgi-
`cal conversion is 80%, but device recovery is only 4% in
`patients who died of non-aneurysm-related causes. Thus it
`is unclear how representative these explant findings are for
`the entire population of implanted devices.
`Structural analysis revealed the presence of one or more
`focal abnormalities in most explanted AneuRx devices,
`including stent fatigue fractures, fabric holes, or suture
`disruption. The structural abnormalities observed were
`few, 2 or 3 stent fractures and fabric holes per device,
`particularly when considered in light of the total number of
`
`structural elements in each device. Thus, for example, the
`number of stent strut fractures found represented less than
`0.2% of the total number of stent struts in a bifurcated
`device. Nonetheless, a relatively small number of stent
`fractures in critical areas of the stent graft may potentially
`have an adverse effect on the structural strength of the
`device. Similarly, even small fabric holes may potentially
`result in type III endoleaks, pressure transmission into the
`aneurysm sac, aneurysm enlargement, and possible aneu-
`rysm rupture.
`To assess the clinical significance of the observed stent
`and fabric abnormalities we reviewed the subset of 37
`clinical trial explants, because these patients, as part of the
`clinical trial, had well-documented clinical follow-up with
`serial imaging studies before explantation. We found no
`difference in the number of fabric holes or fatigue fractures
`in the explanted devices from patients with endoleaks com-
`pared with those without endoleaks before explantation.
`Similarly, no relationship could be established between
`pre-explantation aneurysm enlargement and fabric holes or
`fatigue fractures noted in explanted devices. These obser-
`vations suggest that the structural findings did not account
`for adverse aneurysm-related events such as endoleak and
`enlargement, although the current sample size may not be
`large enough to rule out this possibility. On the other hand,
`
`

`

`8 Zarins et al
`
`JOURNAL OF VASCULAR SURGERY
`July 2004
`
`Fig 6. Explant case example. A, A 24 ⫻ 14 ⫻ 16.5-cm-long bifurcation module was implanted through the right
`femoral artery to treat a 5.5-cm abdominal aortic aneurysm. A 14 ⫻ 8.5-cm-long left iliac limb was used and positioned
`below the bifurcation junction gate (see markers). The left iliac limb was too short, and barely reached the left iliac
`artery. The aneurysm was successfully excluded from the circulation, with no evidence of endoleak (see inserts).
`However, insecure fixation both proximally and distally of the left iliac limb was apparent on the x-ray film. B, Bowing
`of the left iliac limb and slight angulation of the bifurcation module was apparent on x-ray film at 12 months. There
`continued to be no endoleak, and no change in aneurysm size. C, Aneurysm ruptured at 21 months as a result of
`separation of the iliac limb from the bifurcation module. The AneuRx stent graft was explanted during successful
`surgical conversion. D, Explant analysis revealed fatigue fractures in ring 3 of the bifurcation module (top arrows), and
`fabric holes in the angulated right iliac limb at rings 14 to 16, remote from the aneurysm sac (bottom arrows). There
`were no abnormalities in the left iliac limb.
`
`stent-graft migration noted on pre-explantation imaging
`studies was associated with a small increase in both fatigue
`fractures and fabric holes found in explanted devices. It was
`not possible to determine with any degree of certainty
`whether these structural findings developed before or after
`the stent graft migrated.
`Stent fractures found in this study were most com-
`monly located in the bifurcation module, with most frac-
`tures found in aortic rings 3 and 4. At this location the
`round aortic stent ring configuration changes to a bi-oval
`contour to accommodate the two iliac limbs. Stent struts in
`this region are exposed to higher residual strain, and most
`stent strut fractures in this region were individual and
`uninvolved with the fixation seal zone (Fig 2). In one
`instance, five strut fractures were found in aortic stent ring
`3. This was the largest number of fractures found in a single
`stent ring, and was associated with multiple suture breaks.
`This patient had severe aortic neck angulation and eccentric
`compression of the bifurcation module from an angulated
`proximal aortic extender cuff. The bifurcation module sep-
`arated from the aortic extender cuff after 2 years, and the
`patient underwent successful open surgical repair.22 It is
`believed that the angulation stress and motion between the
`proximal extender cuff and bifurcation module led to sep-
`
`aration of the extender cuff.22 The second most common
`site of fatigue fractures was the distal portion of iliac limbs
`in areas of severe angulation and tortuousity. These frac-
`tures were often outside of the aneurysm sac (Fig 6).
`Fabric holes were usually related to fabric abrasion
`against the apex of a stent diamond in areas of severe
`angulation. These focal stent– diamond tip perforations
`were small and were often located in iliac limbs, remote
`from the aneurysm sac. Median hole size was 0.5 mm2,
`approximately the size of a 21-gauge needle. Fabric holes
`observed in explanted devices were likely to have been
`plugged with thrombus or tissue in vivo, and thus it is not
`surprising that there was no significant relationship be-
`tween endoleak and fabric holes. It is possible that such
`sealed holes may account for pressure in the aneurysm sac
`without evidence of endoleak (endotension) and thus lead
`to aneurysm enlargement. However, we found no relation
`between fabric holes and aneurysm enlargement. Aneurysm
`enlargement after endovascular repair in AneuRx clinical
`trial patients, although associated with the presence of an
`endoleak, is not associated with increased risk for aneurysm
`rupture.23 Removal of tissue and thrombus material from
`the stent-graft surface after opening the aneurysm sac dur-
`ing surgical conversion may induce pinpoint bleeding
`
`

`

`JOURNAL OF VASCULAR SURGERY
`Volume 40, Number 1
`
`Zarins et al 9
`
`through openings in the fabric.24 Explant analysis of de-
`vices
`from patients with clinically
`reported “mic-
`roleaks”24,25 revealed that they had received preRPM fabric
`devices, which had weave separation but no evidence of
`fabric holes. No microleaks have been reported in patients
`who have received the AneuRx device made with RPM
`material. Intrinsic polyester fabric failure has been noted in
`aortic grafts used in open surgery,26,27 thus stimulating
`continuing improvement in fabric design. It is likely that
`advanced fabric designs will reduce the risk for abrasion-
`induced holes in the future.
`A comparison of explants removed “for cause” with
`those recovered at “incidental” autopsy revealed a signifi-
`cantly greater number of fatigue fractures and fabric holes
`in “for cause” explants. However, there was also a signifi-
`cant difference in average implant duration between the
`groups, 24 months for “for cause” explants versus only 14
`months for “incidental” explants. Since stent fatigue frac-
`tures and fabric abrasion holes are time-dependent, analysis
`of covariance was used to adjust for the difference in
`duration of implantation. The differences in the number of
`fatigue fractures and fabric holes between groups were not
`significant when adjusted for the differences in impla