Midterm results of endovascular aneurysm sealing to treat abdominal aortic aneurysm Katherine M. Stenson, MD, MRCS, Benjamin O. Patterson, PhD, MRCS, Matthew Joe Grima, MRCS, Jorg L. De Bruin, MS, FRCS, Peter J. E. Holt, PhD, FRCS, and Ian Loftus, MD, FRCS, London, United Kingdom
ABSTRACT Background: Endovascular aneurysm sealing (EVAS) represents a novel approach to the treatment of abdominal aortic aneurysms. It uses polymer technology to achieve an anatomic seal within the sac of the aneurysm. This cohort study reports the early clinical outcomes, technical refinements, and learning curve during the initial EVAS experience at a single institution. Methods: Results from 150 consecutive EVAS cases for intact, infrarenal abdominal aortic aneurysms are reported here. These cases were undertaken between March 2013 and July 2015. Preoperative, perioperative, and postoperative data were collected for each patient prospectively. Results: The median age of the cohort was 76.6 years (interquartile range, 70.2-80.9 years), and 87.3% were male. Median aneurysm diameter was 62.0 mm (IQR, 58.0-69.0 mm). Adverse neck morphology was seen in 69 (46.0%) patients, including aneurysm neck length <10 mm (17.3%), neck diameter >32 mm or <18 mm (8.7%), and neck angulation >60 degrees (15.3%). Median follow-up was 687 days (IQR, 463-897 days); 37 patients (24.7%) underwent reintervention. The rates of unresolved endoleak are 1.3% type IA, 0.7% type IB, and 2.7% type I. There were no type III endoleaks. There have been seven secondary ruptures in this cohort; all but one of these patients survived after reintervention. Only one rupture occurred in an aneurysm that had been treated within the manufacturer’s instructions for use (IFU). Conclusions: The rate of unresolved endoleaks is satisfactorily low. The incidence of secondary rupture is of concern; however, when the IFU are adhered to, the rate is very low. The results of this study suggest that working within the IFU yields better clinical results. (J Vasc Surg 2018;-:1-10.) Keywords: Abdominal aortic aneurysm; Endovascular sealing; Infrarenal
The use of endovascular means to treat abdominal aortic aneurysms (AAAs) has decreased perioperative mortality and hospital stay.1-3 Concerns remain about the durability of these technologies and the need for continuing surveillance to prevent secondary rupture.4-6 It is well known that hostile aortic anatomy, particularly in relation to the aneurysm neck, is associated with increased need for reintervention and poorer outcome.7-13 Endovascular aneurysm sealing (EVAS) with polymer technology was seen as an attractive way of treating a greater proportion of aneurysms within the manufacturer’s instructions for use (IFU).14 We are seeing more aneurysms that do not fit the IFU of contemporary devices, so the boundaries of acceptable practice are
From the St. George’s Vascular Institute, St. George’s Hospital. Author conflict of interest: K.M.S., B.O.P., and P.J.E.H. receive funding from National Institute for Health Research. K.S. and I.L. have provided consultancy services for Endologix. Additional material for this article may be found online at www.jvascsurg.org. Correspondence: Katherine M. Stenson, MD, MRCS, St. George’s Vascular Institute, St. George’s Hospital, Blackshaw Rd, London, SW17 0QT United Kingdom (e-mail:
[email protected]).
being pushed. This device was potentially one with the ability to treat aneurysms not otherwise amenable to endovascular treatment.14 The original IFU (Table I) were broad and potentially included aneurysms with short, wide necks, with very little restriction on the access vessel anatomy. The Nellix device (Endologix Inc, Irvine, Calif) gained its CE mark in 2013. The technique showed promising early results.15-17 The IFU were refined in 2016 (Table I)18 following data from the EVAS FORWARD investigational device exemption trial demonstrating higher than expected rates of distal migration, type IA endoleak, and aneurysm sac expansion. Mathematical modeling was undertaken using these 2-year data to identify the anatomic features that were associated with these complications. As well as addressing both proximal and distal seal zones, the updated IFU address the thrombus burden within the aneurysm by stipulating an aortic lumen ratio, designed to reduce the risk of device migration. This study introduces some longer term follow-up for our well-established EVAS program. It reports on perioperative and postoperative complications and the learning curve and evolution of the technique.
The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any
METHODS
manuscript for which they may have a conflict of interest.
This cohort study reports data collected prospectively from the first 150 consecutive patients to undergo EVAS for intact, infrarenal AAA at this institution between March 2013 and July 2015. Outcomes are reported in
0741-5214 Crown Copyright Ó 2018 Published by Elsevier Inc. on behalf of the Society for Vascular Surgery. All rights reserved. https://doi.org/10.1016/j.jvs.2018.04.016
1
2
Journal of Vascular Surgery
Stenson et al
---
Table I. Instructions for use (IFU): 2013 and 2016 2013 IFU Neck length, mm
$10
Maximum neck diameter, mm
18-32
Neck angulation, degrees
#60
Maximum blood lumen diameter, mm
#70
Iliac diameter, mm
9-35
2016 IFU Neck length, mm
$10
Maximum neck diameter, mm
18-28
Neck angulation, degrees
#60
Maximum blood lumen diameter, mm
#70
Maximum aortic diameter/maximum blood lumen diameter
<1.4
Iliac blood lumen, mm
9-35
Distal seal zone length, mm
$10
Distal seal zone diameter, mm
9-25
accordance with consensus reporting standards issued by the British Society for Endovascular Therapy.19 Advice was taken from the UK National Research Ethics Service, and specific ethical approval was not required as all analyses involve data routinely collected for good clinical care. Patients were counseled about the novel nature of the treatment, and consent was gained for inclusion in the study. Patients underwent EVAS if the anatomy of their aneurysm fell outside the IFU of the conventional endovascular aneurysm repair (EVAR) devices at our disposal but within the broader EVAS IFU. Less than 20% of the patients in this cohort had aneurysms that fell within the IFU of conventional EVAR devices. All aneurysms in this study were treated before the IFU update. All cases were discussed at the weekly aortic multidisciplinary meeting, where treatment plans were agreed on. During the study period, patients continued to be treated at our unit with bifurcated EVAR devices. The first 14 cases reported here fell within the IFU of the Nellix device. This was crucial to allow refinement of the technique. In common with many evolving technologies, initial use was within IFU; but as experience and confidence grew, in conjunction with the early promising results for EVAS, aneurysms with more challenging anatomy were treated with specific consent. Some cases were treated with EVAS instead of fenestrated EVAR or EVAR with parallel grafts. At this time, EVAR with parallel grafts was not licensed for use, and some cases had been turned down by the manufacturer for fenestrated EVAR. Early in the series, EVAS with parallel grafts was not an accepted procedure. During the course of this study, the technique used for EVAS has undergone refinement as experience has grown locally and complications were reported globally.
2018
The current procedure and imaging protocols used at our institution are described in full in the Supplementary Methods (online only). In accordance with consensus reporting standards,19 deaths occurring within 30 days of the EVAS procedure were classified as aneurysm related. Reintervention was considered to be any secondary procedure required to maintain aneurysm exclusion or distal perfusion. Endoleaks were described according to established definitions.20 The IFU issued with the first iteration of the Nellix device in 2013 have been superseded by new IFU issued in 2016; references are made to both of these sets of IFU. Normally distributed continuous variables are expressed as mean 6 standard deviation, and those not normally distributed are expressed as median and interquartile range (IQR). Categorical variables are expressed as number with percentages. Differences between cases treated on or off IFU were assessed using the unpaired t-test, c2 test, and Fisher exact test as appropriate where the data were normally distributed. Where data were not normally distributed, the Mann-Whitney U test was used. A P value <.05 was taken to indicate statistical significance. Kaplan-Meier estimates were used to demonstrate freedom from events, with Breslow and log-rank tests being used to analyze differences between curves. Statistical analysis was undertaken using SPSS (SPSS Inc, Chicago, Ill) and GraphPad (GraphPad Software Inc, La Jolla, Calif).
RESULTS This study reports on 150 consecutive patients who underwent EVAS for intact, infrarenal AAA between March 2013 and July 2015. There were no exclusions. The median age of the patients was 76.6 years (IQR, 70.2-80.9 years), and 12.7% were female. Almost twothirds of the cohort was American Society of Anesthesiologists class 4, reflecting the fact that EVAS was used in some patients who were not deemed fit enough for more complex repairs. All but one case was undertaken under general anesthesia. Baseline characteristics are shown in Table II. Median aneurysm sac diameter was 62.0 mm (IQR, 58.0-69.0 mm), reflecting adherence to international guidelines recommending treatment of aneurysms at a diameter of 55 mm. Sixty-four (42.7%) patients were treated within the 2013 IFU and 23 (15.3%) were treated within the 2016 IFU. The complications observed are summarized in Tables III-VII. There were 22 all-cause deaths during the course of the study, with no statistically significant difference in incidence between those treated on and those treated off IFU. There were four aneurysm-related deaths. Kaplan-Meier estimates for survival are shown in Table III, with no significant differences in those treated on and off IFU. One patient was found collapsed at home on postoperative day 19, having been well on discharge; 1 patient
Journal of Vascular Surgery Volume
-,
Number
Stenson et al
3
-
Table II. Baseline characteristics for all patients Variable Age, years
Overall cohort (N ¼ 150) 75.4 6 7.3
Sex Male
131
Female
19
History of Coronary artery disease
73 (48.7)
Chronic obstructive pulmonary disease
73 (48.7)
Diabetes mellitus
16 (10.7)
Hypertension Cerebrovascular event Smoking
120 (80.0) 28 (18.7) 106 (70.7)
Body mass index, kg/m2
27.7 6 4.7
Serum creatinine concentration, mm
96.1 6 37.8
ASA grade 2
3
3
50
4
97
Anesthesia Regional
1
General
149
Maximum aortic diameter, mm
63.7 6 12.4
ASA, American Society of Anesthesiologists. Categorical variables are presented as number (%). Continuous variables are presented as mean 6 standard deviation.
died of sepsis secondary to an infected common femoral pseudoaneurysm on postoperative day 94; 1 patient, who had been treated with platinum chemotherapy for a bladder neoplasm, underwent an axillobifemoral bypass for bilateral Nellix stent occlusion on postoperative day 159 and died the following day; 1 patient died after a secondary rupture on postoperative day 884, having undergone open surgical repair with explantation of the Nellix system 2 days previously. Reinterventions were significantly more common in those patients treated outside the 2013 IFU (Fig 1), but no such difference was seen with the 2016 IFU, and no difference was noted with increasing experience. Kaplan-Meier estimates of freedom from reintervention are 84.4%, 79.9%, and 70.5% at 1 year, 2 years, and 3 years, respectively. Seventeen type IA endoleaks (Fig 2) were reported, two of which spontaneously resolved. Eleven occurred in aneurysms that were treated outside the IFU because of adverse neck morphology and were seen predominantly in the early part of our experience, although these findings are not statistically significant. Kaplan-Meier estimates of freedom from type IA endoleak are 94.4%, 92.9%, and 90.1% at 1 year, 2 years, and 3 years. We have previously reported some issues with proximal positioning of the stent grafts, leading to low deployment,21
caused by insufficient correction of parallax in angulated necks. Subsequently, the procedure was modified to allow correction of parallax using the sheath markers of the Nellix system itself to ensure positioning of the X-ray tube perpendicular to the aneurysm neck. In some early cases, the Nellix stents appeared to deviate from their deployment position on inflation of the endobags. This seemed more common in large aneurysms with angulated necks. The stents were seen to separate and therefore move laterally in a large sac, thus displacing the stents distally. This led to a modification in technique whereby the Nellix balloons were kept inflated during the prefill and polymer curing, not only to prevent compression but also to stabilize stent position. Nine cases were treated with embolization using Onyx (ev3, Irvine, Calif) with a scaffold of coils (Fig 3), three underwent proximal extension using either Atrium (Maquet Cardiovascular, Hudson, NH) or Fluency (C.R. Bard Inc, Tempe, Ariz) stents, and three were treated with a Nellix-in-Nellix procedure (Table VI). Of these reinterventions, one required another intervention, that being repeated embolization, which was successful. One type IA endoleak resulted in secondary rupture, and the patient underwent open surgical repair with explantation of the Nellix system, which he survived. Reintervention for type IA endoleak took place at a median of 294 days (IQR, 21-847 days; Table V). Type IB endoleaks were less common and again occurred early in the experience. Kaplan-Meier estimates of freedom from type IB endoleak are 98.6% at 1 year, 2 years, and 3 years. Four have been reported. Three patients had relining and extension of the affected limb with another Nellix stent; one of these cases also required embolization of the ipsilateral internal iliac artery. One distal endoleak was not reintervened on because of the increasing medical frailty of the patient and the lack of sac size increase. The first generation of the device did not feature apposition of the endobags to the distal ends of the stents. This meant that on occasion, the endobag filling did not proceed down into the iliac system sufficiently to prevent retrograde filling of the aneurysm sac. A later iteration of the device, in use from early 2015, obviated this particular problem. Type IB endoleak was implicated in one case of sac expansion, which led to secondary rupture, successfully treated with distal extension of the device. The median time for reintervention for type IB endoleak was 564 days (IQR, 91-1134 days), suggesting that this form of leak becomes apparent later than is the case for type IA and type II. Seven type II endoleaks were reported. Kaplan-Meier estimates of freedom from type II endoleak are 98.0%, 97.1%, and 94.4% at 1 year, 2 years, and 3 years. Four were not associated with sac size increase and have not required any reintervention. Of those requiring treatment, two underwent embolization of the inferior mesenteric artery and a lumbar artery, respectively; the
4
Journal of Vascular Surgery
Stenson et al
---
2018
Table III. Kaplan-Meier estimates for freedom from complications, including comparisons with respect to compliance with instructions for use (IFU) and case number Kaplan-Meier survival estimate, %
Log-rank test, P value Case number (1-50, 51-100, 101-150)
1 Year
2 Years
3 Years
On vs off 2013 IFU
On vs off 2016 IFU
Type IA endoleak
94.4
92.9
90.1
.591
.219
.471
Type IB endoleak
98.6
98.6
98.6
.092
.639
.604
All type I endoleaks
93.1
91.5
88.7
.834
.342
.493
Type II endoleak
98.0
97.1
94.4
.175
.918
.482
Migration
99.2
98.3
89.5
.026
.118
.416
Reintervention
84.4
79.9
70.5
.155
.185
.557
All-cause mortality
93.3
86.3
82.3
.266
.412
.164
Aneurysm-related mortality
98.0
98.0
96.5
.447
.380
.352
Complication
Boldface P values indicate statistical significance (P < .05).
Table IV. Reinterventions for endoleaks Complication Type IA endoleak
Type IB endoleak
Type II endoleak
No. 17
4
7
Reintervention
Table V. Median times for reintervention No.
Complication
Onyx embolization
9
Type IA endoleak
Nellix-in-Nellix
3
Type IB endoleak
564
Proximal extension
3
Type II endoleak
246
OSR and explantation of Nellix for rupture
1
Migration
865
Sac size increase
564
No reintervention
2
Limb occlusion
204
Limb relining/extension
3
Rupture
584
Embolization of IIA
1
No reintervention
1
Embolization of IMA
1
Embolization of lumbar artery
1
Nellix-in-Nellix for rupture
1
No reintervention
1
IIA, Internal iliac artery; IMA, inferior mesenteric artery; OSR, open surgical repair.
other was implicated in a secondary rupture, which was successfully treated with a Nellix-in-Nellix procedure. It is thought that the type II endoleak led to a sac expansion, in turn causing the proximal seal to deteriorate and allow a proximal endoleak. Type II endoleaks were treated at a median of 246 days (IQR, 153-1075 days). Stent migration is taken to refer to distal displacement of the proximal part of the stent in relation to the lowermost renal artery by >5 mm. Kaplan-Meier estimates of freedom from migration are 99.2%, 98.3%, and 89.5% at 1 year, 2 years, and 3 years. Migration was seen to be significantly more common when treatment was outside the 2013 IFU (P ¼ .026; Fig 4). Eight migrations were reported; five were associated with a type IA endoleak, and of these, two have seen an increase in sac size. Both cases were treated successfully with a Nellix-inNellix procedure. The other three cases resulting in type
Median time of reintervention, days 294
IA endoleak were treated effectively with proximal extension (two cases) and embolization of the leak. One case of migration resulted in secondary rupture; this aneurysm had a particularly adverse neck in that it was short and angulated. The patient was treated with an open repair and explantation of the Nellix, but he died 2 days postoperatively. Migration appears to be a relatively late finding and has been treated at a median of 865 days (IQR, 352-1197 days); as such, it is possible that as more patients in this cohort progress in their followup, more cases of migration will become apparent. Expansion of the aneurysm sac (>5 mm in maximum diameter) was seen in 11 cases. Three such cases were associated with type IA endoleak and were treated with a Nellix-in-Nellix procedure. Two were associated with type II endoleaks and one with a type IB endoleak and were treated with embolization of the inferior mesenteric, lumbar, and internal iliac arteries. Four sac size increases were not associated with an obvious endoleak; these remain asymptomatic. Median time for intervention was 564 days (IQR, 153-1197 days). Nellix limb occlusion was observed on seven occasions. Four occurred within the first 50 cases and in retrospect were probably the result of inadequate stent dilation during the original procedure and placement of the limb at a site of a potential kink. As previously described,
Journal of Vascular Surgery Volume
-,
Number
Stenson et al
5
-
Table VI. Complications broken down by number of 2013 instructions for use (IFU) parameters breached Overall
0 Parameters (n ¼ 65)
1 Parameter (n ¼ 55)
2 Parameters (n ¼ 20)
3 Parameters (n ¼ 4)
Type IA endoleak
17
6 (9.2)
8 (14.5)
2 (10.0)
1 (25.0)
Type IB endoleak
4
3 (4.6)
0
0
1 (25.0)
.017
Type II endoleak
5
1 (1.5)
4 (7.3)
0
0
.269
Migration
8
2 (3.1)
3 (5.5)
3 (15.0)
0
.225
Sac size increase
9
5 (7.7)
4 (7.3)
0
0
.590
Limb occlusion
7
4 (6.2)
1 (1.8)
2 (10.0)
0
.446
Secondary rupture
7
4 (6.2)
2 (3.6)
1 (5.0)
0
.893
All-cause mortality
22
7 (10.8)
7 (12.7)
8 (40.0)
0
.010
Aneurysm-related mortality Reintervention
P value .675
4
1 (1.5)
0
3 (15.0)
0
.005
34
14 (21.5)
13 (23.6)
6 (30.0)
1 (25.0)
.986
Values are reported as number (%). Boldface P values indicate statistical significance (P < .05).
Table VII. Complications broken down by number of 2016 instructions for use (IFU) parameters breached Overall
0 Parameters (n ¼ 23)
1 Parameter (n ¼ 44)
2 Parameters (n ¼ 42)
3 Parameters (n ¼ 20)
4 Parameters (n ¼ 12)
P 5 Parameters (n ¼ 3) value
Type IA endoleak
17
1 (4.3)
6 (13.6)
5 (11.9)
4 (20.0)
1 (8.3)
0
Type IB endoleak
4
1 (4.3)
2 (4.5)
0
0
0
1 (33.3)
.022
Type II endoleak
5
1 (4.3)
0
3 (7.1)
1 (5.0)
0
0
.553
Migration
8
0
2 (4.5)
4 (9.5)
1 (5.0)
1 (8.3)
0
.693
Sac size increase
9
1 (4.3)
3 (6.8)
5 (11.9)
0
0
0
.446
Limb occlusion
6
1 (4.3)
1 (2.3)
2 (4.5)
2 (10.0)
0
0
.661
Secondary rupture
7
1 (4.3)
3 (6.8)
0
2 (10.0)
1 (8.3)
0
.535
All-cause mortality
22
2 (8.7)
5 (11.4)
8 (18.2)
4 (20.0)
2 (16.7)
1 (33.3)
.723
Aneurysm-related mortality
4
0
1 (2.3)
1 (2.3)
2 (10.0)
0
0
.418
34
4 (17.4)
13 (29.5)
7 (15.9)
7 (35.0)
3 (25.0)
0
.438
Reintervention
.672
Values are reported as number (%). Boldface P values indicate statistical significance (P < .05).
Fig 1. Reintervention-free survival with respect to compliance with 2013 instructions for use (IFU).
the procedure has been refined such that the Nellix balloons remain inflated during the curing process, and postdilation is routinely undertaken. In addition, in cases in which a kink is considered likely, a self-expanding Zilver stent (Cook Medical, Bloomington, Ind) is placed distally to reduce the risk of a flow-limiting lesion. These improvements seem to have had a beneficial effect
on the incidence of limb occlusions, with three reports in cases 51 to 150, one being in a patient with an acquired prothrombotic state secondary to treatment with platinum-containing chemotherapy. Reinterventions for limb occlusion were the earliest of all complications, at a median of 204 days (IQR, 159-319 days). It would appear that the decrease in incidence as time has progressed is a true representation of an improvement in technique rather than any lag effect. Two occlusions were treated with thrombectomy of the Nellix limb and relining, one with a femorofemoral crossover graft and one with an axillobifemoral graft. Of the seven ruptures that occurred, all were reintervened on and six of the patients survived. These interventions occurred at a median of 584 days (IQR, 500-884 days) and represented four open repairs with explantation of the device, two Nellix-in-Nellix procedures, and one distal limb extension with embolization of the internal iliac artery. All but one of the patients in this group had had unremarkable surveillance scans up until the point of rupture. One patient had been noted to have a small
6
Stenson et al
Journal of Vascular Surgery ---
2018
Fig 2. Coronal computed tomography image of a type IA endoleak after endovascular aneurysm sealing (EVAS).
increase in sac size, but there was no apparent endoleak on either computed tomography or duplex ultrasound scanning. The computed tomography angiograms taken on presentation with the rupture revealed 2 type IA endoleaks, 1 type IB endoleak with sac expansion, 1 type II endoleak, and 1 migration. There was no significant difference in the rates of rupture between those cases treated on and those cases treated off IFU. Freedom from rupture is shown in Fig 5. Reference is made to the original 2013 IFU and the refined 2016 IFU. “Inside” and “outside” IFU are rather blunt terms; some morphologic features may breach limits by one parameter, others by five. Tables VI and VII break the complications down by the number of IFU parameters that are breached for each set of IFU. For both sets of IFU, there was a statistically significant rise in the number of type IB endoleaks reported as more IFU parameters were breached (P ¼ .017 for the 2013 IFU and .022 for the 2016 IFU). Regarding the 2013 IFU, both all-cause mortality and aneurysm-related mortality were significantly higher in cases in which more parameters were breached (P ¼ .009 and .004, respectively). Kaplan-Meier charts showing freedom from mortality according to the number of IFU parameters breached are shown in Figs 6 and 7.
DISCUSSION New medical technologies are inevitably associated with a range of complications that require definition and management. The cases presented here represent
Fig 3. Coronal computed tomography image of a type IA endoleak treated with Onyx and coils.
the first group of patients treated with this novel system at a single institution. When this clinical program was initiated, the global experience with this device was limited. As a result, during the time frame of this study, the technical aspects of the procedure have evolved and become refined. Our experience and the global experience have led to a better understanding of which aneurysms should be treated with the Nellix device. Adverse proximal neck anatomy remains the Achilles heel for endovascular aneurysm treatment. Previous studies have shown us that adverse neck anatomy is associated with poorer outcomes and a greater need for reintervention.22-24 When the Nellix graft was first made commercially available, it was hoped that it might be able to treat a greater proportion of aneurysms with morphologic features within the manufacturer’s IFU, particularly with regard to the aneurysm neck. It was thought that the polymer sealing technology would be able to create an adequate seal even within short, conical, and angulated necks. In retrospective analysis, almost half of this cohort (69 patients) have aneurysms with adverse neck morphology. As confidence grew with this new device and with promising early results, more patients with challenging morphologic features were treated. Thus, more aneurysms were treated outside the IFU. These were often patients for whom there was no other endovascular option.
Journal of Vascular Surgery Volume
-,
Number
Stenson et al
7
-
Fig 4. Migration-free survival with respect to compliance with 2013 instructions for use (IFU).
Fig 5. Rupture-free survival with respect to case number.
Fig 6. All-cause mortality-free survival with respect to number of 2013 instructions for use (IFU) parameters breached.
Some of the early type IA endoleaks seen in this series were undoubtedly the result of technical inexperience, and these lessons have been learned and the procedure refined accordingly. This particularly relates to the adequate correction of parallax to avoid low placement of the proximal ends of the Nellix stents. This is of paramount importance as it ensures that there is sufficient effacement of the endobag to healthy aorta to avoid endoleak. In addition, the maneuver whereby the Nellix
Fig 7. Aneurysm-related mortality-free survival with respect to number of 2013 instructions for use (IFU) parameters breached.
balloons are kept inflated during primary fill and polymer curing has been developed during the course of this study to allow stabilization of the stents, thus avoiding the relative graft shortening produced by the bowing of stents within a large aneurysm sac with an angled neck. However, not all of the type IA endoleaks are related to the learning curve with the new device. Higher incidences of these complications have been reported than with some other EVAR devices, and this may also relate to the very challenging morphologic features that have been treated in this series. The rate of secondary rupture in this study is 2.58 per 100 patient-years, occurring at a median of 584 days postoperatively. This compares with early and late rupture rates of 7.2 and 0.6 per 100 patient-days, respectively, as reported in the UK EVAR trials.25 Another concern about such a new technology is the potential difficulty in treating new complications or recognizing familiar complications in a new setting. The appearance of the Nellix stent graft system is different from that of conventional bifurcated devices. The polymer within the endobags is initially radiodense, potentially making endoleaks occurring within 3 months difficult to detect. The location of endoleaks after EVAS can be different compared with EVAR; for example, type IA endoleaks may be observed at the peripheries of the aneurysm sac (between the endobag and the aneurysm sac) or even in the cleft in between the endobags.26-28 Treating type IA endoleaks after EVAS is a different proposition from treating them after EVAR. In our series, we have used three methods: embolization with Onyx and coils; extension of the proximal stents individually; and more recently, extension of the sealing zone proximally with another pair of Nellix stents inserted through the original pair.29 Embolization using Onyx, an ethylene vinyl alcohol copolymer, instilled through a microcatheter placed in the proximal neck, has gained some popularity, and some small published series have shown good results.30,31 Our experience has shown that distal type I endoleaks are less common but occur later in follow-up than do
8
Journal of Vascular Surgery
Stenson et al
---
types IA and II. The second-generation Nellix device, which came into use in early 2015, appears to have decreased the incidence of type IB endoleaks. The endobags in the earlier version were not attached to the stents at the distal ends; this did allow retrograde filling of the aneurysm sac from the iliac arteries. The newer devices have addressed this apparent design flaw, and we will determine the success of this when the later cases in the series have longer term follow-up. One aim of the EVAS device was to reduce the risk of type II endoleak. This complication may still occur when the endobags are not able to fully unfurl, for example, when the blood flow lumen is narrow secondary to high thrombus burden, thus allowing a small pocket of potential space, which may allow a type II endoleak if it is in communication with a branch vessel. In this study, 4.7% of cases saw a type II endoleak, compared with 16.8% and 14.6% at midterm follow-up using older (pre-2004) and newer stent grafts as reported in a single-center study of 1412 patients.32 High thrombus volumes within the aneurysm sac have been reported as a concern globally. Despite few publications on this matter, there is evidence to suggest that the thrombus volume within the aneurysm sac changes considerably after EVAS and may have implications for device performance.33 Collected anecdotal evidence suggests that low-volume retrograde flow from the iliac vessels, through an inadequate distal seal, may potentially contribute to an increase in thrombus volume. The potential problems surrounding proximal and distal sealing and intraluminal thrombus in the sac led to a tightening of the IFU in 2016. The accepted luminal diameters both proximally and distally have been narrowed, and a greater emphasis has been placed on distal sealing zone. This has come from the observation that even when the distal sac is sealed, if the distal end of the stent is not in apposition with the iliac wall, retrograde flow may push the endobag proximally with formation of a type IB endoleak. This may particularly be the case in aortoiliac aneurysms, in which backflow from the internal iliac artery may lead to sac size increase if the common iliac artery is not completely excluded.34 An appreciation of the thrombus load is also encouraged by the updated IFU by calculating the ratio between the maximum aortic and flow lumen diameters. The rates of migration, type IA endoleak, and secondary rupture are higher than reported in other EVAR studies; however, the rate of type II endoleak is lower. Type IA endoleak certainly exists as a mode of failure for this device, particularly in conjunction with migration, and may lead to secondary rupture. Retrospective morphologic analysis of the aneurysms before and after treatment has shown the importance of placing the stents at the level of the renal arteries to maximize the aortic neck. It seems that migration occurs more commonly when this is not the case and when EVAS is used to treat
2018
short aneurysm necks; indeed, the Kaplan-Meier estimates suggest significantly greater migration-free survival in aneurysms treated on IFU (Table III). There is a significantly higher rate of all-cause mortality when more 2013 IFU parameters are breached. It has previously been shown that more complex aneurysms are associated with a greater risk of mortality secondary to multisystem atherosclerosis, whether or not the aneurysm is treated.35,36 As a single-center study, we cannot be sure how well these data will extrapolate to those from other institutions. Results have been presented comparing compliance and noncompliance with both sets of IFU; for the 2016 IFU, the small number of aneurysms treated within IFU makes is difficult to draw any statistically significant conclusions. Other centers are building sizable cohorts of patients who have been treated with EVAS, and so it will be fascinating to compare experiences and outcomes. The incidence of complications is more common in our earlier experience, but other than for all-cause mortality, these differences are not statistically significant. This may be related to improving technique, but it may also be because later in our experience, more of the complex morphologic features were treated with chimney grafts. These data report on our earliest experiences with this device and so detail the technical refinements that have been made, with special attention to proximal deployment of the stents, increased ballooning during polymer cure, and more widespread use of adjunctive iliac stenting to avoid limb occlusions. Although this is a series of consecutive patients treated with EVAS for intact, infrarenal aneurysms, during the course of the study, other EVAR grafts were being used at the same institution. This means that there does remain the potential for selection bias to have taken place.
CONCLUSIONS This study reports on an elderly cohort of patients with challenging aortic anatomy, some of whom do not fit within the IFU for any commercially available endovascular device. Early procedure-related issues were resolved with refinement of the technique and, when combined with increasing experience, yielded improving patient outcomes from a clinical perspective, albeit not statistically significant. These data do provide evidence for the safety and effectiveness of the Nellix system; however, it remains crucially important to operate within the IFU where possible. Although the difference in incidence of complications within and without IFU is not statistically significant, the absolute numbers are small and are clinically significant. Medium-to long-term outcomes, especially with respect to aorta-related complications and their reinterventions, will determine the widespread applicability and durability of this novel treatment.
Journal of Vascular Surgery Volume
-,
Number
Stenson et al
9
-
AUTHOR CONTRIBUTIONS Conception and design: KS, BP, MG, JDB, PH, IL Analysis and interpretation: KS, BP, MG, JDB, PH, IL Data collection: KS, JDB Writing the article: KS Critical revision of the article: KS, BP, MG, JDB, PH, IL Final approval of the article: KS, BP, MG, JDB, PH, IL Statistical analysis: KS Obtained funding: Not applicable Overall responsibility: KS
16.
REFERENCES
17.
1. Lederle FA, Freischlag JA, Kyriakides TC, Padberg FT Jr, Matsumura JS, Kohler TR, et al. Outcomes following endovascular vs open repair of abdominal aortic aneurysm: a randomized trial. JAMA 2009;302:1535-42. 2. Prinssen M, Verhoeven EL, Buth J, Cuypers PW, van Sambeek MR, Balm R, et al. A randomized trial comparing conventional and endovascular repair of abdominal aortic aneurysms. N Engl J Med 2004;351:1607-18. 3. Greenhalgh RM, Brown LC, Kwong GP, Powell JT, Thompson SG; EVAR trial participants. Comparison of endovascular aneurysm repair with open repair in patients with abdominal aortic aneurysm (EVAR trial 1), 30-day operative mortality results: randomised controlled trial. Lancet 2004;364:843-8. 4. United Kingdom EVAR Trial Investigators, Greenhalgh RM, Brown LC, Powell JT, Thompson SG, Epstein D, Sculpher MJ, et al. Endovascular versus open repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1863-71. 5. De Bruin JL, Baas AF, Buth J, Prinssen M, Verhoeven EL, Cuypers PW, et al. Long-term outcome of open or endovascular repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1881-9. 6. Lederle FA, Freischlag JA, Kyriakides TC, Matsumura JS, Padberg FT Jr, Kohler TR, et al. Long-term comparison of endovascular and open repair of abdominal aortic aneurysm. N Engl J Med 2012;367:1988-97. 7. Patterson BO, Hinchliffe RJ, Holt PJ, Loftus IM, Thompson MM. Importance of aortic morphology in planning aortic interventions. J Endovasc Ther 2010;17:73-7. 8. Karthikesalingam A, Holt PJ, Vidal-Diez A, Choke EC, Patterson BO, Thompson LJ, et al. Predicting aortic complications after endovascular aneurysm repair. Br J Surg 2013;100:1302-11. 9. Schanzer A, Greenberg RK, Hevelone N, Robinson WP, Eslami MH, Goldberg RJ, et al. Predictors of abdominal aortic aneurysm sac enlargement after endovascular repair. Circulation 2011;123:2848-55. 10. Aburahma AF, Campbell JE, Mousa AY, Hass SM, Stone PA, Jain A, et al. Clinical outcomes for hostile versus favorable aortic neck anatomy in endovascular aortic aneurysm repair using modular devices. J Vasc Surg 2011;54:13-21. 11. Antoniou GA, Georgiadis GS, Antoniou SA, Kuhan G, Murray D. A meta-analysis of outcomes of endovascular abdominal aortic aneurysm repair in patients with hostile and friendly neck anatomy. J Vasc Surg 2013;57:527-38. 12. El Batti S, Cochennec F, Roudot-Thoraval F, Becquemin JP. Type II endoleaks after endovascular repair of abdominal aortic aneurysm are not always a benign condition. J Vasc Surg 2013;57:1291-7. 13. Cieri E, De Rango P, Isernia G, Simonte G, Ciucci A, Parlani G, et al. Type II endoleak is an enigmatic and unpredictable
14.
15.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
marker of worse outcome after endovascular aneurysm repair. J Vasc Surg 2014;59:930-7. Karthikesalingam A, Cobb RJ, Khoury A, Choke EC, Sayers RD, Holt PJ, et al. The morphological applicability of a novel endovascular aneurysm sealing (EVAS) system (Nellix) in patients with abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2013;46:440-5. Donayre CE, Zarins CK, Krievins DK, Holden A, Hill A, Calderas C, et al. Initial clinical experience with a sacanchoring endoprosthesis for aortic aneurysm repair. J Vasc Surg 2011;53:574-82. Krievins DK, Holden A, Savlovskis J, Calderas C, Donayre CE, Moll FL, et al. EVAR using the Nellix sac-anchoring endoprosthesis: treatment of favourable and adverse anatomy. Eur J Vasc Endovasc Surg 2011;42:38-46. Brownrigg JR, de Bruin JL, Rossi L, Karthikesalingam A, Patterson B, Holt PJ, et al. Endovascular aneurysm sealing for infrarenal abdominal aortic aneurysms: 30-day outcomes of 105 patients in a single centre. Eur J Vasc Endovasc Surg 2015;50:157-64. Reijnen M, Holden A. Status of endovascular aneurysm sealing after 5 years of commercial use. J Endovasc Ther 2018;25:201-6. Boyle JR, Thompson MM, Vallabhaneni SR, Bell RE, Brennan JA, Browne TF, et al. Pragmatic minimum reporting standards for endovascular abdominal aortic aneurysm repair. J Endovasc Ther 2011;18:263-71. White GH, Yu W, May J, Chaufour X, Stephen MS. Endoleak as a complication of endoluminal grafting of abdominal aortic aneurysms: classification, incidence, diagnosis, and management. J Endovasc Surg 1997;4:152-68. de Bruin JL, Brownrigg JR, Karthikesalingam A, Patterson BO, Holt PJ, Hinchliffe RJ, et al. Endovascular aneurysm sealing for the treatment of ruptured abdominal aortic aneurysms. J Endovasc Ther 2015;22:283-7. Jones JE, Atkins MD, Brewster DC, Chung TK, Kwolek CJ, LaMuraglia GM, et al. Persistent type 2 endoleak after endovascular repair of abdominal aortic aneurysm is associated with adverse late outcomes. J Vasc Surg 2007;46:1-8. Fulton JJ, Farber MA, Sanchez LA, Godshall CJ, Marston WA, Mendes R, et al. Effect of challenging neck anatomy on midterm migration rates in AneuRx endografts. J Vasc Surg 2006;44:932-7; discussion: 937. Abbruzzese TA, Kwolek CJ, Brewster DC, Chung TK, Kang J, Conrad MF, et al. Outcomes following endovascular abdominal aortic aneurysm repair (EVAR): an anatomic and device-specific analysis. J Vasc Surg 2008;48:19-28. Wyss TR, Brown LC, Powell JT, Greenhalgh RM. Rate and predictability of graft rupture after endovascular and open abdominal aortic aneurysm repair: data from the EVAR trials. Ann Surg 2010;252:805-12. Karthikesalingam A, de Bruin JL, Patel SR, Azhar B, Rossi L, Morgan RA, et al. Appearance of the Nellix endovascular aneurysm sealing system on computed tomography: implications for postoperative imaging surveillance. J Endovasc Ther 2015;22:297-302. Hilliard NJ, Hayes P, Winterbottom AP. Appearance of the Nellix endoprosthesis on postoperative imaging: implications for patient and device surveillance. Semin Vasc Surg 2016;29:126-34. D’Abate F, Harrison SA, Karthikesalingam A, de Bruin J, Hinchliffe RJ, Loftus IM, et al. Sonographic appearance following endovascular aneurysm repair using the Nellix endovascular sealing system. J Endovasc Ther 2015;22: 182-6.
10
Stenson et al
Journal of Vascular Surgery ---
29. Donselaar EJ, Holden A, Zoethout AC, Zeebregts CJ, Reijnen MM. Feasibility and technical aspects of proximal Nellix-in-Nellix extension for late caudal endograft migration. J Endovasc Ther 2017;24:210-7. 30. Chun JY, Morgan R. Transcatheter embolisation of type 1 endoleaks after endovascular aortic aneurysm repair with Onyx: when no other treatment option is feasible. Eur J Vasc Endovasc Surg 2013;45:141-4. 31. Henrikson O, Roos H, Falkenberg M. Ethylene vinyl alcohol copolymer (Onyx) to seal type 1 endoleak. A new technique. Vascular 2011;19:77-81. 32. Verzini F, Isernia G, De Rango P, Simonte G, Parlani G, Loschi D, et al. Abdominal aortic endografting beyond the trials: a 15-year single-center experience comparing newer to older generation stent-grafts. J Endovasc Ther 2014;21: 439-47. 33. Shaikh U, Chan TY, Oshin O, McWilliams RG, Fisher RK, England A, et al. Changes in aortic volumes following endovascular sealing of abdominal aortic aneurysms with the Nellix endoprosthesis. J Endovasc Ther 2015;22:881-5.
2018
34. Krievins DK, Savlovskis J, Holden AH, Kisis K, Hill AA, Gedins M, et al. Preservation of hypogastric flow and control of iliac aneurysm size in the treatment of aortoiliac aneurysms using the Nellix EndoVascular Aneurysm Sealing endograft. J Vasc Surg 2016;64:1262-9. 35. Bahia SS, Holt PJ, Jackson D, Patterson BO, Hinchliffe RJ, Thompson MM, et al. Systematic review and meta-analysis of long-term survival after elective infrarenal abdominal aortic aneurysm repair 1969-2011: 5 year survival remains poor despite advances in medical care and treatment strategies. Eur J Vasc Endovasc Surg 2015;50:320-30. 36. Forsdahl SH, Solberg S, Singh K, Jacobsen BK. Abdominal aortic aneurysms, or a relatively large diameter of nonaneurysmal aortas, increase total and cardiovascular mortality: the Tromso study. Int J Epidemiol 2010;39:225-32. Submitted Nov 10, 2017; accepted Apr 10, 2018.
Additional material for this article may be found online at www.jvascsurg.org.
Journal of Vascular Surgery Volume
-,
Number
Stenson et al
10.e1
-
SUPPLEMENTARY METHODS (online only). Detailed operative and surveillance protocol. During the course of this study, the technique used for endovascular aneurysm sealing has undergone refinement as experience has grown locally and complications were reported globally. The device comprises two cobaltchromium stents coated with expanded polytetrafluoroethylene, each surrounded by a polyurethane “endobag,” into which is instilled a biocompatible polyethylene glycol-based polymer. This polymer is initially an aqueous solution, which then cures in 3 to 5 minutes. The current procedure used at our institution is described here. Preoperative aortic morphology was accurately assessed with three-dimensional reconstructions using 3mensio software (3mensio Medical Imaging BV, Bilthoven, The Netherlands). The most important decisions to be made at this stage are the likely device length that will be required and the estimated polymer fill volume; the device length is calculated by measuring the centerline distance from the renal arteries to the iliac bifurcation, and the fill volume is calculated by the flow volume between the renal arteries and the aortic bifurcation. The common femoral arteries are punctured percutaneously under ultrasound guidance, and two ProGlide closure devices (Abbott Vascular, Abbott Park, Ill) are deployed in each artery. Unfractionated heparin is administered intravenously at this stage, usually at a dose of 5000 units but dependent on the patient’s body mass; 14F sheaths (Cook Medical, Bloomington, Ind) are introduced over Lunderquist wires (Cook Medical), and a calibrated angiogram is taken to confirm the Nellix (Endologix Inc, Irvine, Calif) stent length that is required. Subsequently, the Nellix devices themselves are inserted over the wire, through the common femoral arteries bilaterally, into a position at the level of the ostium of the lowest renal artery. Once parallax has been corrected using the sheath markers, an angiogram is taken through the angiogram ports in the device’s nose cone. After retraction of the covering sheath, the stents are placed at the level of the lowest renal artery and then deployed simultaneously with inflation of the Nellix balloons to 7 atm. The balloons then remain inflated during the polymer instillation and curing process to ensure that the stents do not become compressed and that the endograft system as a whole maintains maximum stability. Once the stents have been deployed, the endobags surrounding them are “prefilled” with saline to a pressure of 180 mm Hg to confirm the sac volume estimations.
An angiogram is obtained here also to ensure that sealing at this defined pressure and volume will avoid endoleak. Should an endoleak be apparent at this time, further saline can be instilled into the endobags to a maximum pressure of 220 mm Hg. Once adequate sealing has been demonstrated, the saline is aspirated from the endobags and is replaced with an equal volume of aqueous polyethylene glycol-based polymer, which is left to cure for up to 5 minutes. Completion of the curing process is clear from the appearance of the residual polymer in the fill lines. Another angiogram is taken to ascertain sealing of the aneurysm sac. At this stage, the primary fill lines are removed, and pressure within the sac is checked using the secondary fill lines and a small volume of saline. Should there be either a failure of adequate sealing or an endobag pressure of <180 mm Hg, more polymer can be introduced as a “secondary fill.” “Postdilation” of the Nellix stents is undertaken routinely using a 10- 40-mm angioplasty balloon (Cook Medical) inflated to 7 atm. Completion angiography is employed to confirm complete sealing of the sac and patency of the flow channels. In situations in which there is unacceptable angulation between the stent graft and the wall of the iliac vessel, an additional uncovered, bare-metal self-expandable stent (Zilver; Cook Medical) is deployed. All patients are prescribed a single antiplatelet agent unless there is a specific contraindication. Duplex ultrasound scanning is carried out by specialist vascular sonographers, who use a Philips iU22 machine (Philips Healthcare, Eindhoven, The Netherlands) and a C5-1 MHz curvilinear array. The ultrasound surveillance protocol is designed to interrogate graft patency, aneurysm sac dimension change, presence of endoleak, and complications that may arise at the percutaneous access site, such as pseudoaneurysm. The computed tomography (CT) angiography protocol produces images from the thoracic outlet to the bifurcation of the common femoral artery, with a slice thickness of 0.625 mm. The GE Lightspeed VCT 64 Slice Scanner (GE Healthcare, Waukesha, Wisc) is used along with 90 mL of intravenous Omnipaque 300 (iohexol 300/Omnipaque 300; Sanofi-Winthrop, New York, NY), according to departmental protocol. A consensus on the required surveillance after endovascular aneurysm sealing has yet to be achieved. Our patients routinely have a duplex ultrasound examination and CT angiography imaging before discharge. Subsequent to this, they undergo surveillance with both duplex ultrasound and CT at 6 and 12 months and annually thereafter.