Editor
’s Choice e Occurrence and Classification of Proximal Type I Endoleaks
After EndoVascular Aneurysm Sealing Using the Nellix
Ô Device
Leo H. van den Hama, Andrew Holdenb, Janis Savlovskisc, Andrew Witterbottomd, Kenneth Ouriele, Michel M.P.J. Reijnena,*, on behalf of the EVAS Type IA Endoleak Study Group
a
Department of Surgery, Department of Radiology, Rijnstate Hospital, Arnhem, The Netherlands
b
Department of Radiology, Auckland Hospital, Auckland, New Zealand
c
Department of Radiology, Stradins University Hospital, Riga, Latvia
d
Department of Radiology, Addenbrooke’s Hospital, Cambridge, UK
e
Syntactx, New York, NY, USA
WHAT THIS PAPER ADDS
Endovascular aneurysm sealing (EVAS), using the Nellix device, is a novel approach to treat infrarenal aneu-rysms. The incidence and types of proximal endoleaks are still poorly understood. This study contributes to the knowledge on the incidence and types of proximal type I endoleaks after EVAS. In addition, it describes potential causes and treatment options.
Objective/Background:Proximal type I endoleaks are associated with abdominal aortic aneurysm (AAA) growth and rupture and necessitate repair. The NellixÔ EndoVascular Aneurysm Sealing (EVAS) system is a unique approach to AAA repair, where the appearance and treatment of endoleaks is also different. This study aimed to analyse and categorise proximal endoleaks in an EVAS treated cohort.
Methods:All patients, treated from February 2013 to December 2015, in 15 experienced EVAS centres, presenting with proximal endoleak were included. Computed tomography scans were analysed by a core laboratory. A consensus meeting was organised to discuss and qualify each case for selection, technical aspects, and possible causes of the endoleak. Endoleaks were classified using a novel classification system for EVAS.
Results:During the study period 1851 patients were treated using EVAS at 15 centres and followed for a median of 494 283 days. Among these, 58 cases (3.1%) developed a proximal endoleak (1.5% early and 1.7% late); of these, 84% of 58 patients were treated outside the original and 96% outside the current, refined, instructions for use. Low stent positioning was the most likely cause in 44.6%, a hostile anatomy in 16.1%, and a combination of both in 33.9%. Treatment, by embolisation or proximal extension, was performed in 47% of cases, with a technical success of 97%.
Conclusion:The overall incidence of proximal endoleak after EVAS is 3.1% after a mean follow-up period of 16 months, with 1.5% occurring within 30 days. Their occurrence is related to patient selection and stent
positioning. Early detection and classification is crucial to avoid the potential of sac rupture. Ó 2017 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved.
Article history: Received 8 May 2017, Accepted 25 September 2017, Available online 28 October 2017
Keywords:Classification, EndoVascular aneurysm sealing, Endoleak, EVAS, Incidence, Nellix
INTRODUCTION
Type I endoleaks after endovascular aneurysm repair (EVAR) are related to aneurysm growth and rupture and usually require treatment.1,2 EndoVascular Aneurysm Sealing (EVAS), using the NellixÔ system (Endologix, Irvine, CA, USA), is another method for the treatment of infrarenal abdominal aortic aneurysms (AAAs).3,4 The technique is based on two balloon expandable covered stents sur-rounded by endobags that arefilled in situ with a soluble
polymer. These polymerfilled endobags provide both fixa-tion and seal. The technique was designed in an attempt to reduce complications, particularly endoleaks of any kind, and subsequent re-interventions during follow-up. The in-structions for use (IFU) at its introduction in 2013 included an infrarenal neck diameter of 18e32 mm, a neck length of 10 mm, and a
b
angle < 60. Recently, the IFU has been redefined to further optimise outcome with a reduction of neck diameter range to a maximum of 28 mm, reduction of* Corresponding author. Department of Surgery, Rijnstate Hospital, Wagnerlaan 55, 6815 AD Arnhem, The Netherlands. E-mail address:mmpj.reijnen@gmail.com(Michel M.P.J. Reijnen).
1078-5884/Ó 2017 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. https://doi.org/10.1016/j.ejvs.2017.09.016
the maximum iliac artery diameter to 20 mm, and the addition of an AAA/lumen ratio (> 1.4). In addition, a second generation was introduced in 2014 with, amongst other improvements, distalfixation of the endobags to the stents. In early publications EVAS was used not only inside, but also outside the IFU.5 EVAS has also been used in combination with parallel grafts to treat more complex anatomy and for failed EVAR.6e8The published incidence of endoleak after EVAS is low. The Nellix system Investigational Device Exemption (IDE) pivotal trial, with 142 patients treated within the IFU, reported a total endoleak rate at 30 days of 6.3% (type I, 0.7%; type II, 5.6%).9 At 1 year the persistent endoleak rate was 3.1% (type I, 0.8%; type II, 2.3%). The EVAS FORWARD Global Registry, which enrolled 277 patients both inside and outside the IFU, had an early type Ia endoleak in eight cases.10Root cause analysis sug-gested inadequate use of the proximal seal zone and under filling of the endobags to be the most likely cause. The KaplaneMeier estimate of freedom from type I endoleak at 12 month follow-up was 96%. The published incidence of endoleak is low in the short-term, but there are increasing concerns about the durability of the device.
The diagnosis of a type Ia endoleak after EVAS can be challenging as its appearance is different from EVAR. The radio density of the polymer within endobags is usually high early after implantation, rendering the early diagnosis difficult.11Optimal treatment modalities have also yet to be defined, although embolisation and proximal extension techniques have been suggested. The current study was conducted in order to gain more knowledge of the inci-dence and types of type Ia endoleak after EVAS, to assess the potential causes, and to study the treatment options.
METHODS
The design was a retrospective, multicentre, observational study, and all sites that had treated> 50 patients with EVAS at January 2016 were invited to participate. Patients from 15 participating sites were included from February 2013, when the device received its CE mark, until March 2016. The participating centres all performed at least yearly computed tomography (CT) scans, with the additional use of duplex ultrasound. Data were evaluated at 30 days and most recent follow-up. The study was approved by each centre’s ethical committee, according to the national guidelines.
Procedural aspects of EVAS have been described previ-ously.3,4 All patients treated by EVAS, both inside and outside the IFU, and presenting with type Ia endoleak, either early or late, were included. Procedural type Ia endoleaks that were detected and treated during the initial procedure (i.e., with secondaryfill) were excluded, as were patients treated by EVAS in conjunction with chimney grafts and revision EVAR cases. Clinical data were retrieved from the medical records and included demographics, comor-bidities, procedural details, and follow-up. All CT angiog-raphy (CTA) studies were analysed by a core laboratory (Syntactx, New York, NY, USA) and a consensus meeting took place on 23 May 2016 with four EVAS experienced
physicians in attendance (AH, JS, AW, MMPJR). All available imaging of the patients was anonymised and studied, spe-cifically looking at patient selection, technical aspects, diagnosis, and treatment of the endoleak.
Definitions and endpoints
AAAs were classified as inside or outside the original and refined IFU. Early type Ia endoleak was defined as a prox-imal endoleak, detected on either the completion angio-gram, within 30 days on any imaging modality or on thefirst preformed post-procedural imaging.12 Late endoleak was defined as occurring after the first post-procedural imaging study had been normal.
Comorbidity was scored according to the Society for Vascular Surgery and American Association for Vascular Surgery medical comorbidity scoring system. Complications within 30 days of the index procedure were considered to be procedure related. Aneurysm related re-intervention was defined when a secondary procedure was performed to address a complication related to the device, the index procedure, or to the aneurysm itself. The seal length was calculated by deducting the length from the lowest renal artery to the top of the stent from the original neck length and deducting another 5 mm as this is where the endobags start.
Post-EVAS type Ia endoleaks were categorized as follows: a type Is1 endoleak was defined as the appearance of contrast between the endobag and the wall of the proximal neck but not reaching the aneurysm sac. This type would not be classified as endoleak within the accepted definitions for EVAR. Type Is2 endoleaks were defined as those where there was contrast between the endobag and aneurysmal wall or thrombus inside the aneurysm sac; a type Is3 endoleak was defined as showing contrast or newly formed thrombus between the endobags inside the aneurysm sac; and a type Is4 endoleak was defined as the presence of sac pressurisation without proof of endoleak or with the pres-ence of secondary signs (Fig. 1).
Statistical analysis
All statistical analyses were performed using SPSS Statistics version 20.0 (IBM, Armonk, NY, USA). Continuous variables are presented as mean SD, and categorical variables as n (%). Intergroup analysis was done using the ManneWhitney U test for continuous variables, and the chi-square test and Fisher’s exact test for categorical variables. Two sided p values of < .05 were considered statistically significant.
RESULTS
During the study period 1851 EVAS procedures were per-formed in the participating sites, by 71 vascular surgeons, which was about one third of the overall number of EVAS procedures performed worldwide at that time. Sixty-six cases with a proximal type I endoleak were identified by the sites. After core laboratory analysis eight were excluded from further analysis owing to the absence of a proximal type I endoleak (n¼ 4), revision EVAR cases (n ¼ 2), and a
procedural endoleak, which was treated intra-operatively (n ¼ 2). As such, the final study group consisted of 58 cases, with a median follow-up after EVAS of 501 (inter-quartile range [IQR] 251e699) days. Baseline patient char-acteristics are displayed inTable 1.
EVAS procedure
Fifty-three patients (91%) were treated in an elective setting and 45 (78%) under general anaesthesia. The mean procedural time was 123 55 min, with a mean blood loss of 229 261 mL. Pre-fill was performed at a mean pressure of 184 8 mmHg, with a volume of 112 86 mL saline. A combination of saline and contrast medium was used as the pre-fill solution in 24 (41%) cases. All patients were treated with the first generation Nellix EVAS (3SQ). The mean polymer volume was 118 84 mL and in 43/58 patients (19%) post-ballooning was applied. In five (9%) cases a secondaryfill was performed with a mean of 9 11 mL of polymer. Fourteen (24%) concomitant procedures were performed: 10 iliac extensions, two angioplasties, and two thrombo-endarterectomies.
Type Ia endoleak diagnosis and treatment
The overall incidence of type Ia endoleak was 3%. The median interval between treatment and detection was 157 days (IQR 382). Twenty-seven cases (1.5%) were early and the remaining 31 cases (1.7%) were late. Thirteen (22%) of the late type Ia endoleaks were discovered within 1 year, and four (6%) within 2 years. The majority (n¼ 48, 83%) of
type Ia endoleaks were detected on CTA (two ECG gated), eight on duplex ultrasound (14%), and two on angiography (3%).
Core laboratory assessment
Follow-up imaging was available in 56 of 58 cases, but one of them was excluded because of inadequate baseline im-aging quality. Baseline anatomy is displayed inTable 2. After EVAS the maximum diameter of the aneurysm sac increased from 65.0 13.2 to 67.2 13.1 mm (p < .001). There was a straightening of the infrarenal neck angulation (p¼ .007) and the angle between the AAA and right common iliac artery (p< .001), which was observed over a one month time span from implantation tofirst follow-up. In the early endoleak group the infrarenal neck length was 14.8 10.8 mm, of which 6.0 0.9 mm was used for seal. In the late endoleak group the neck length was 15.1 9.7 mm, of which 10.6 1.7 mm seal length was achieved. At the time of endoleak diagnosis this seal length was reduced to 7.3 0.6 mm.
Eighty-four percent of the study group was treated outside the original IFU (Table 3) with 96% outside the refined IFU (Table 4). The percentage of patients treated outside the IFU in the rest of the population was unknown In 21 of 56 cases (37%) the endoleak started as an Is1 endoleak. Of these, six (28%) and three (14%) developed into an Is2 and Is3 endoleak during follow-up. Twenty-three of 56 cases (41%) presented as an Is2 endoleak, of which one progressed into an Is3. Eight of 56 cases (14%) pre-sented as an Is3 and four of 56 cases (7%) as an Is4, of which one developed into an Is3 endoleak.
Overall, three patients presented with an AAA sac rupture, of which two were initially treated outside the original and the refined IFUs.
Consensus meeting
In 17 cases (29%) early signs of the endoleak were identified on imaging that was performed earlier than the reported time of diagnosis (i.e., endobag separation, contrast in the Figure 1. (A) Type Is1 endoleak defined as the appearance of contrast between the endobag and the wall of the proximal neck but not reaching the aneurysm sac itself. (B) Type Is2 endoleak defined as the appearance of contrast between the endobag and aneurysmal wall or thrombus inside the aneurysm sac. (C) Type Is3 was defined as the appearance of contrast or fresh thrombus between the endobags inside the aneurysm sac. (D) Type Is4 endoleak defined as the presence of sac pressurisation without proof or with the presence of secondary signs.
Table 1. Baseline demographics (n¼ 58).
Male sex 43 (74)
Mean SD age (y) 79 8
Hypertension 47 (81) Cardiac history 31 (53) Pulmonary history 44 (76) History of stroke 8 (14) Diabetes mellitus 8 (14) Smoking (current) 11 (19)
Note. Data are n (%) unless otherwise indicated.
Table 2. Anatomical data.
All p Early EL p Late EL p
Pre-operative 30 day Pre-operative 30 day Pre-operative 30 day At time of
endoleak Suprarenal neck diameter at
SMA (mm)
26.0 3.0 26.4 3.0 .072 26.1 3.1 26.5 2.8 .208 25.9 3.0 26.3 3.2 26.4 3.6 .093
Neck diameter at lowest RA (mm)
24.7 3.5 25.6 3.6 .002 24.2 2.9 25.1 3.1 .071 25.1 4.0 26.1 4.0 26.5 4.4 .349
Neck diameter 10 mm below lowest RA
26.6 5.4 29.1 5.4 < .001 26.4 4.7 28.9 5.6 .013 26.9 6.1 29.2 5.3 29.7 5.8 .112
Infrarenal neck length (mm) 14.9 10.1 e e 14.8 10.8 e e 15.1 9.7 e e e
Range 2.0e44.4 3.1e44.4 2.0e37.4
Infrarenal neck lumen volume (mL)
6.9 12.1 e e 8.5 17.1 e e 5.4 3.9 e e e
Total seal length (mm) e 8.3 0.8 e e 6.0 0.9 e e 10.6 1.7 7.3 0.6 e
Infrarenal neck angulation (40 mm rule)
29.0 14.0 23.0 13.0 .007 31.0 15.0 26.0 13.0 .048 26.0 13.0 21.0 31.0 23 10.0 .380
Maximum AAA sac diameter (mm)
65.0 13.2 67.2 13.1 < .001 64.7 13.3 67.2 13.3 .011 65.2 13.3 67.3 13.2 70.8 13.4 .750
Aortic diameter at bifurcation (mm)
23.2 5.3 e e 22.3 5.2 e e 24.1 5.4 e e e
Max AAA cross sectional area (mm2)
31.2 14.4 33.5 14.0 < .001 31.1 14.1 33.3 13.9 .009 31.3 14.8 33.8 14.4 32.6 13.8 .898
Aortic sac volume (outer wall) (mL)
232.7 149.1 251.7 147.1 < .001 229.2 147.1 247.7 150.3 .006 235.9 153.6 255.2 146.8 243.2 127.6 .738
Aorticflow lumen volume (mL) 110.0 85.4 e e 111.9 72.9 e e 108.1 97.3 e e e
Length from lowest RA to aortic bifurcation (mm)
128.2 22.5 131.8 20.4 .036 126.6 23.3 128.1 20.7 .337 129.8 22.1 135.1 20.0 131.0 18.3 .150
Right CIA length (mm) 58.1 18.1 52.1 18.5 < .001 60.7 18.0 53.4 15.3 .012 55.7 18.1 51.0 21.0 54.9 17.6 .312
Max right CIA diameter (mm) 19.8 6.8 20.7 6.1 < .001 20.2 5.8 20.6 4.6 .001 19.5 7.6 20.7 7.2 22.3 7.1 .067
Min right EIA luminal diameter (mm)
7.9 1.7 e e 7.6 1.7 e e 8.1 1.7 e e e
Aorta right CIA angulation (o) 34.0 22.0 24.0 14.0 < .001 34.0 20.0 25.0 17.0 .001 34.0 24.0 24.0 10.0 27.0 15.5 .491
Left CIA length (mm) 58.1 19.3 52.1 17.5 < .001 58.5 20.3 51.5 18.6 < .001 57.7 18.6 52.7 16.7 51.9 16.5 .512
Max left CIA diameter (mm) 18.3 5.2 19.2 5.4 <0.001 17.8 4.1 19.0 4.7 0.013 18.7 (6.1) 19.4 6.1 19.7 5.7 0.326
Min left EIA luminal diameter (mm)
7.8 1.8 e e 7.5 1.7 e e 8.0 1.9 e e e
Aorta left CIA angulation (o) 41.0 21.0 30.0 16.0 .812 44.0 23.0 31.0 17.0 .001 39.0 18.0 29.0 14.0 28.66 16.2 .177
Note. Data are mean SD. EL ¼ endoleak; SMA ¼ superior mesenteric artery; RA ¼ renal artery; AAA ¼ abdominal aortic aneurysm; CIA ¼ common iliac artery; EIA ¼ external iliac artery.
Leo H. van den Ham et al.
infrarenal neck between endobag and wall) and as such were considered a delayed or initially missed diagnosis.
Every endoleak except one had one of the following: hostile anatomy for EVAS (short and/or angulated necks, “stomach shaped” AAA) in 16.1%; low and/or misaligned stents in 44.6%; and a combination of both in 33.9%. Low stent placement was seen in 28 patients (48%), underfilling of both endobags in two patients (3%) and an offset posi-tion of the stents in five patients (9%). Distal migration
during follow-up was found in 10 (17%) cases, of which two resulted in a rupture (Fig. 2).
Re-intervention options were discussed per case. In 10 cases (17%) embolisation with coils and Onyx/glue was considered to be appropriate. In 42 cases (72%) proximal Nellix-in-Nellix extension, with the use of chimneys, was considered the optimal re-intervention. In the remaining six (10%) patients, conversion to open repair was considered to be best, owing to a progressive endoleak, development into the Is2/Is3 subtype, and an unsuitable anatomy for proximal extensions.
Treatment
After detection of the proximal type I endoleak, a re-intervention was performed in 40 cases (69%). Three other endoleaks which were all categorized as type Is1, were reported as “resolved” and did not return during follow-up. Core laboratory analysis categorized these pa-tients as still having a type Is1 endoleak. Thirteen papa-tients (22%) were converted to open repair,five (9%) were treated using coil embolisation only, 12 (21%) patients underwent a combination of coils and Onyx or glue embolisation, and 10 (17%) patients were treated using proximal Nellix-in-Nellix extensions with chimneys. The overall technical success of the re-interventions was 96.5%, with one patient under-going conversion surgery after embolisation complications using Onyx. One patient died 19 days after conversion to open surgery owing to cerebral haemorrhage. Follow-up of patients treated by coil embolisation was short and varied between 1 and 6 months.
Of the untreated patients two were scheduled for re-intervention and six were still in conservative follow-up. One patient was left untreated owing to a malignancy and two other patients were considered too frail for con-version and died owing to unknown causes. Three patients presented with rupture and died on, respectively, day 42, day 204, and day 393 after EVAS. All three presented with Table 3. Outside original instructions for use (IFU) (46/55 patients;
84%). Neck length (< 10 mm)a 10 (18) Neck diameter (< 18 mm >32 mm) 2 (4) < 18 mm 1 (2) > 32 mm 1 (2) Neck angulation (> 60) 7 (13) Sac lumen (> 60 mm) 3 (5)
Iliac artery diameter (< 9 mm > 35 mm) 42 (76)
< 9 mm 41 (75)
> 35 mm 2 (4)
Note. Data are n (%).aDefinition for location of distal end of neck changed from 10% increase from the diameter of the lowest renal in the original IFU to 20% in the refined IFU.
Table 4. Outside refined instructions for use (53/55 patients; 96%). Neck length (< 10 mm)a 20 (36) Neck diameter (< 18 mm > 28 mm) 10 (18) < 18 mm 1 (2) > 28 mm 9 (16) Neck angulation (> 60) 7 (13) Sac lumen (> 60 mm) 3 (6) AAA/lumen ratio ( 1.4) 31 (56)
Iliac artery diameter 48 (87)
< 9 mm 42 (76)
> 20 mm 23 (42)
Note. Data are n (%). AAA ¼ abdominal aortic aneurysm.
aDefinition for location of distal end of neck changed from 10%
increase from the diameter of the lowest renal in the original IFU to 20% in the refined IFU.
Figure 2. Computed tomography demonstrating stent migration; positioning at (A) 1 month and (B) 3 years after EndoVascular Aneurysm Sealing, showing a 4 mm distal migration of both stents without visible endoleak.
an Is3 endoleak. The last patient died unrelated to AAA before re-intervention
DISCUSSION
In the current study, it has been shown that proximal type I endoleaks occur after EVAS with an overall incidence of 3% in short-term follow-up, which is in line with the frequency reported for EVAR. The occurrence of proximal type I endoleak appeared to be related to patient selection and technical aspects of the procedure. Over 85% of patients with proximal type I endoleak were treated outside the original IFU, mostly owing to neck length, neck angulation, and iliac artery diameter. Whereas thefirst two are common risk factors for type Ia endoleak, small common iliac arteries may also complicate EVAS. Placement of balloon expand-able stents in narrow common iliac arteries limits reposi-tioning options and as such could affect proximal positioning. The role of patient selection was also clear as in 50% of cases the anatomy was considered to be hostile and in 34% this was combined with suboptimal positioning. After the introduction of EVAS, a perception existed that sealing of the aneurysm sac alone would be sufficient and that patients unsuitable for EVAR could be treated with EVAS. This could explain the observation that many patients with a hostile anatomy were treated with this novel tech-nique and that, in patients with a suitable anatomy, stents were deliberately placed low in the infrarenal neck. This is reflected in the observation that, particularly in the early endoleaks, only a small part of the infrarenal neck was used for the seal. The calculated sealing length is probably even an overestimation of seal as it suggests an optimal circumferential wall apposition of the endobags, which is often not the case. Further studies on this, studying the endobag position instead of the stent position, are indi-cated but challenging. With growing experience, better patient selection, and technique, the incidence of type Ia endoleak after EVAS might further decrease.
Current best practice consists of patient selection within the IFU and optimal stent positioning, with the upper margin of the endobagsflush below the most caudal renal artery. The necessity of complete use of the infrarenal neck is nowadays appreciated, but it was not in the early phase after introduction of the device. Accurate placement can be very challenging and as a consequence low placement could be unintentional. The use of a pre-fill of the endobags with saline solution is advised to verify seal with 180 mmHg endobag pressure. Leaving the Nellix balloons up during pre-fill and polymer fill may facilitate maintenance of stent position, especially in difficult anatomy. It is advised to perform an angiogram in multiple views through a pigtail catheter after one Nellix device has been removed, allowing secondaryfill through the other device in case of insuffi-cient seal.
Distal migration was the commonest cause of late endoleaks, as reflected by the decreased seal length over time. When positioned off-set or low, especially in a short neck, any migration could immediately compromise
proximal seal. The incidence of migration is largely un-known, although a> 4 mm displacement was described in five of 18 EVAS treated patients.13At 1 year, core laboratory adjudicated migration >10 mm occurred in three patients (2.3%) in the Nellix system IDE pivotal trial.9Migration was not reported in the 12 month outcome of the EVAS FOR-WARD Global Registry,10 but this was only site reported. Downward force on the polymer shelf and acceleration forces within curvatures of the stents might both contribute. In the authors’ experience, caudal migration is often related to a low polymer volume in relation to high thrombus volume. This observation underlines the addition of the maximum aneurysm to maximum luminal diameter to the revised IFU. It also emphasizes the potential for unexpected complications in new concepts and devices and underlines that they should always be introduced with utmost care.
The diagnosis of type Ia endoleaks after EVAS can be difficult and differs from EVAR, as previously described.11In the present series, the endoleaks could have been detected at an earlier stage in 29% of cases. It was also shown that they tend to progress over time into more severe subtypes. Early detection and treatment is therefore essential and CT follow-up is the preferred modality. Based on the current data the value of plain radiography and duplex ultrasound is not clear. The analysis prompted the design of a novel classification system with implications for clinical practice.
All observed early endoleaks in the current study were either Is1 or Is2. Based on the present observations, type Is1 is prone to progression and as such is clinically relevant. For that reason they were classified as endoleaks in this study, whereas they would not have been following EVAR. This obviously complicates comparisons with the incidence of EVAR. This morphology reduces seal and the cause is likely to be related to patient selection and/or underfilling of endobags. Presence of thrombus in the aortic neck in continuity with the aneurysmal sac could lead to pressure transmitted to the aneurysmal sac. Patients should be un-der close observation and in case of progression treatment is advocated.
Type Is2 was observed both early and late and is likely to cause pressurisation of the aneurysmal sac at the area of the endoleak. This type may have a relationship with an outflow vessel. Although they are less likely to disappear spontaneously, type Is2 endoleaks with outflow tend to be more benign with regard to sac enlargement than cases without. In the latter situation, similar to dissections, the end-diastolic pressure will be high owing to a lack of outflow, raising the mean arterial pressure within the endoleak thus increasing the risk of complications.14 The inflow section of the endoleak could act as a one way ball valve and in this case sac pressure would rise to the highest pressure over a period of time. The occurrence of the type Is2 endoleaks was related to procedural aspects, patient selection, and, in late failures, distal migration.
Type Is3 endoleaks were observed only during follow-up and should be considered to be the most aggressive, as the reconstruction is unstable. This type of endoleak will
pressurize the entire aneurysm sac and causes lateral dislocation of both stents, with direct loss of positional stability and endobag wall apposition. It is closely related to migration and was observed in all secondary ruptures with reported endoleak.
Radiological appearances suggestive of a loss of seal included: (i) progression of aneurysm diameter or volume; (ii) progression of aneurysm thrombus volume; (iii) newly formed thrombus between the endobags; (iv) endobags undergoing morphological changes; (v) progressive bowing of one or both stents. These signs are suggestive of an occult T1 endoleak or “endotension (Is4)” and are con-cerning, especially when related to a short seal zone. Additional imaging modalities, such as contrast enhanced ultrasound, electrocardiogram gated CTA, and magnetic resonance angiography, could be considered.
Treatment options for proximal type I endoleaks after EVAS are limited to embolisation or proximal extension using a Nellix-in-Nellix application with/without parallel grafts. A significant number of patients were converted to open in this cohort. Complications after EVAS may require another approach to EVAR and the decision to open con-version may be taken earlier. The latter is particularly true in the early phase when treatment options for complications, including embolisation and Nellix-in-Nellix extension, still needed to be explored. Treatment of type Is1 endoleak should aim to improve the seal and the gutter could be obliterated using coils in combination with Onyx or glue,15e17although durability of these treatments remains to be shown. Type Is2 endoleaks require treatment and options include embolisation with coils and Onyx/glue when the position of the stents is considered optimal. For distal migration or malpositioning a proximal Nellix-in Nellix extension is indicated to achieve adequate seal.18 These new Nellix devices should extend above the original Nellix stents for at least 30 mm. In case of chimney cases, a seal zone of 20 mm is advocated. The endovascular treatment option for type Is3 endoleak is a proximal Nellix-in Nellix extension with/without chimney grafts in order to re-establish a seal. In Is4 endoleaks careful observation is advised or treatment of the most likely cause in cases of progression. Open surgical repair may be necessary, particularly in type Is2 and Is3.
The current study has limitations. Owing to the retro-spective research design not all imaging was available and minor, subclinical type Ia endoleaks could have been missed leading to under reporting. The identification of patients was left to the participating centre’s discretion, which could also lead to under reporting. The detection of endoleaks is dependent on each centre’s follow-up scheme and use of different imaging modalities. In addition, the follow-up period is still relatively short and therefore the data should be interpreted with care. However, imaging of 56 of the included patients was complete and independently analysed by a core laboratory and judged by a group of Nellix EVAS experienced radiologists. Treatment options have only limited follow-up and sample sizes are small. The
durability of the various treatment modalities therefore requires further research.
CONCLUSION
The overall incidence of type Ia endoleak is 3.1% after a mean follow-up of 16 months, with 1.5% occurring within 30 days. Their occurrence appears related to patient se-lection and stent positioning. Early detection is crucial as they tend to develop into more severe types and there is a need for a new classification system, as has been described.
CONFLICT OF INTEREST
This study was sponsored by Endologix. Michel Reijnen, Andrew Holden, and Janis Savlovskis are consultants for Endologix, and have received funding for research and speaker’s fees.
FUNDING
This study was sponsored by Endologix.
APPENDIX 1. Collaborators and institutions involved in the EVAS Type IA Endoleak Study Group.
Contributor Site
Leo van den Ham Michel Reijnen
Department of Vascular Surgery, Rijnstate Hospital, Arnhem, The Netherlands
Andrew Holden Department of Interventional Radiology, Auckland Hospital, Auckland, New Zealand Janis Savlovskis
Dainis Krievins
Department of Radiology Department of Surgery, Pauls Stradins Clinical University Hospital, Riga, Latvia Andrew Winterbottom
Paul Hayes
Department of Radiology Department of Vascular Surgery, Cambridge University Hospitals NHS Trust, Cambridge, UK Jan Heyligers Department of Vascular Surgery,
St. Elisabeth Hospital, Tilburg, The Netherlands
Dittmar Böckler Department of Vascular and Endovascular Surgery, Heidelberg University Hospital, Heidelberg, Germany
Jean-Paul de Vries Department of Vascular Surgery, St. Antonius Hospital,
Nieuwegein, The Netherlands Sebastian Zerwes
Rudi Jakob
Department of Vascular Surgery, Hospital of Augsburg, Augsburg, Germany
Marwan Youssef Department of Cardiothoracic and Vascular Surgery, Medical Centre of the Johannes Gutenberg-University Mainz, Germany
Patrick Berg Roland Stroetges
Department of Vascular Surgery, Marienhospital Kevelaer, Kevelaer, Germany
Continued
REFERENCES
1 Moll FL, Powell JT, Fraedrich G, Verzini F, Haulon S, Waltham M, et al. Management of abdominal aortic aneurysms clinical practice guidelines of the European society for vascular sur-gery. Eur J Vasc Endovasc Surg 2011;41:S1e58.
2 Tan TW, Eslami M, Rybin D, Doros G, Zhang WW, Farber A. Outcomes of patients with type I endoleak at completion of endovascular abdominal aneurysm repair. J Vasc Surg 2016;63: 1420e7.
3 van den Ham LH, Zeebregts CJ, de Vries JP, Reijnen MM. Abdominal aortic aneurysm repair using NellixTMEndoVascular Aneurysm Sealing. Surg Technol Int 2015;26:226e31. 4 Krievins DK, Holden A, Savlovskis J, Calderas C, Conayre CE,
Moll FL, et al. EVAR using the Nellix Sac-anchoring endopros-thesis: treatment of favourable and adverse anatomy. Eur J Vasc Endovasc Surg 2011;42:38e46.
5 Böckler D, Reijnen MM, Krievins D, Peters AS, Hayes P, De Vries JP. Use of the Nellix EVAS system to treat post-EVAR complications and to treat challenging infrarenal necks. J Cardiovasc Surg (Torino) 2014;55:601e12.
6 Dijkstra ML, Lardenoye JW, van Oostayen JA, Zeebregts CJ, Reijnen MM. Endovascular aneurysm sealing for juxtarenal aneurysm using the Nellix device and chimney covered stents. J Endovasc Ther 2014;21:541e7.
7 Stella N, Ficarelli R, Dito R, Brancadoro D, Rossi M, Taurino M, et al. A double Nellix and chimney covered stents: challenging
treatment of pararenal aortic aneurysm. Vasc Endovascular Surg 2017;51:209e14.
8 van den Ham LH, Wiersema SM, Kievit JK, Reijnen MM. Treat-ment of type IIIb endoleaks after EVAR using the Nellix EndoVascular Aneurysm Sealing System. J Endovasc Ther 2016;23:29e32.
9 Carpenter JP, Cuff R, Buckley C, Healey C, Hussain S, Reijnen MM, et al. One-year pivotal trial outcomes of the Nellix system for endovascular aneurysm sealing. J Vasc Surg 2017;65. 330e336 e4.
10 Thompson MM, Heyligers JM, Hayes PD, Reijnen MM, Böcker D, Schelzig H, et al. Endovascular aneurysm sealing: early and midterm results from the EVAS FORWARD Global Registry. J Endovasc Ther 2016;23:685e92.
11 Holden A, Savlovskis J, Winterbottom A, van den Ham LH, Hill A, Krievins D, et al. Imaging after Nellix Endovascular Aneurysm Sealing: a consensus document. J Endovasc Ther 2016;23:7e20. 12 Chaikof EL, Blankensteijn JD, Harris PL, White GH, Zarins GK, Bernhard VM, et al. Reporting standards for endovascular aortic aneurysm repair. J Vasc Surg 2002;35:1048e60. 13 England A, Torella F, Fisher RK, McWilliams RG. Migration of
the Nellix endoprosthesis. J Vasc Surg 2016;64:306e12. 14 Tsai TT, Schlicht MS, Khanafer K, Bull JL, Valassis DT,
Williams DM, et al. Tear size and location impacts false lumen pressure in an ex vivo model of chronic type B aortic dissec-tion. J Vasc Surg 2008;47:844e51.
15 Ameli-Renani S, Morgan RA. Transcatheter embolisation of proximal type 1 endoleaks following Endovascular Aneurysm Sealing (EVAS) using the Nellix device: technique and out-comes. Cardiovasc Intervent Radiol 2015;38:1137e42. 16 Eberhardt KM, Sadeghi-Azandaryani M, Worlicek S, Koeppel T,
Reiser MD, Treitl M. Treatment of type I endoleaks using transcatheter embolization with onyx. J Endovasc Ther 2014;21:162e71.
17 Ameli-Renani S, Pavlidis V, Morgan RA. Early and midterm outcomes after transcatheter embolization of type I endoleaks in 25 patients. J Vasc Surg 2017;65:346e55.
18 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;24:210e7.
APPENDIX 1-continued
Contributor Site
Alexander Oberhuber Department of Vascular and Endovascular Surgery, University of Düsseldorf, Düsseldorf, Germany
Eric Zimmermann Department of Radiology, SIHF, Hamar, Norway
Michelangelo Ferri Department of Vascular Surgery, Umberto I Hospital, Turin, Italy Jorg de Bruin St George’s Vascular Institute, St
