- Email: [email protected]
Changes in Renal Anatomy After Fenestrated Endovascular Aneurysm Repair
Eur J Vasc Endovasc Surg (2017) 53, 95e102
Changes in Renal Anatomy After Fenestrated Endovascular Aneurysm Repair5 B. Maurel a,b, Y. Lounes a, M. Amako a, D. Fabre a, A. Hertault a, J. Sobocinski a, R. Spear a, R. Azzaoui a, T.M. Mastracci b, S. Haulon a,* a b
Aortic Center, Hôpital Cardiologique, CHRU de Lille, France Aortic Team, Royal Free London NHS Foundation Trust, London, UK
WHAT THIS PAPER ADDS Fenestrated devices have become the standard of care for juxtarenal aneurysms. Knowing the impact that variability in renal stent position has on long-term outcomes may influence clinical pathways. This study explores the changing positions of renal anatomy after fenestrated repair.
Objective: To assess short- and long-term movement of renal arteries after fenestrated endovascular aortic repair (FEVAR). Methods: Consecutive patients who underwent FEVAR at one institution with a custom-made device designed with fenestrations for the superior mesenteric (SMA) and renal arteries, a millimetric computed tomography angiography (CTA), and a minimum of 2 years’ follow-up were included. Angulation between renal artery trunk and aorta, clock position of the origin of the renal arteries, distance between renal arteries and SMA, and target vessel occlusion were retrospectively collected and compared between the pre-operative, post-operative (<6 months), and last (>12 months) CTA. Results: From October 2004 to January 2014, 100 patients met the inclusion criteria and 86% of imaging was available for accurate analysis. Median follow-up was 27.3 months (22.7e50.1). There were no renal occlusions. A significant change was found in the value of renal trunk angulation of both renal arteries on post-operative compared with pre-operative CTA (17 difference upward [7.5e29], p < .001), but no significant change thereafter (p ¼ .5). Regarding renal clock positions (7.5 of change equivalent to 15 min of renal ostial movement): significant anterior change was found between post-operative and pre-operative CTA (15 min [0e 30], p ¼ .03 on the left and 15 min [15e30], p < .001 on the right), without significant change thereafter (15 min [0e30], p ¼ .18 on the left and 15 min [0e15] on the right, p ¼ .28). No changes were noted on the distance between renal and SMA ostia (difference of 1.65 mm [1e2.5], p ¼ .63). Conclusion: The renal arteries demonstrate tolerance to permanent changes in angulation after FEVAR of approximately 17 upward trunk movement and of 15e30 min ostial movement without adverse consequences on patency after a median of more than 2 years’ follow-up. The distance between the target vessels remained stable over time. These results may suggest accommodation to sizing errors and thus a compliance with off the shelf devices in favourable anatomies. Ó 2016 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. Article history: Received 18 July 2016, Accepted 24 October 2016, Available online 24 November 2016 Keywords: FEVAR, Renal angulation, Renal clock position, Sizing, Off the shelf devices
INTRODUCTION Over the past two decades, endovascular repair of abdominal aortic aneurysms (EVAR) has gained popularity as a treatment option that offers many advantages over 5 This study has been presented in the plenary session at the ESVS 30th Annual Meeting in Copenhagen, Denmark, on 28e30 September 2016 (abstract n 1357). * Corresponding author. Aortic Centre, CHRU de Lille, INSERM U1008, Université Lille Nord de France, 59037 Lille Cedex, France. E-mail address: [email protected] (S. Haulon). 1078-5884/Ó 2016 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejvs.2016.10.019
open surgical repair. However, specified anatomical characteristics are required to achieve durable outcomes, and up to 40% of patients are unsuitable for traditional infrarenal EVAR, mainly because of a hostile proximal neck anatomy.1,2 To overcome these limitations, use of a fenestrated stent graft for fenestrated endovascular aneurysm repair (FEVAR) currently represents the most validated and reliable endovascular option to treat short neck or pararenal aneurysms, with high technical success rates (99%), low operative mortality (3.5%), and excellent mid- and long-term target vessel patency (97%).3,4 Fenestrated stent grafts are custom-made to fit each patient’s anatomy, implying a delay for planning,
96
production, and implantation from 6 to 8 weeks. The success of the procedure relies on appropriate sizing of the stentgraft, including accurate determination of target vessel clock face position, distance from proximal stentgraft edge, and aortic diameter, to ensure aneurysm exclusion and target vessel patency. Sizing is commonly performed by the manufacturer in a centralised planning centre. If experienced surgeons do their own planning, it is a complex process that includes learning anatomical limitations to meet manufacturing constraints and requires a 3-dimensional workstation. Moreover, concerns have been raised over intra- and inter-observer variability and potential consequences.5e7 As not all patients are able to wait for a customised graft, off the shelf fenestrated devices, including three fenestrations for the superior mesenteric artery (SMA) and renal arteries, and a scallop for the coeliac trunk (CT), available for immediate use and designed to accommodate the majority of the patients are under investigation with encouraging early results.8 It has been shown to be applicable in 70% of the patients, with the right renal artery as a primary cause of exclusion. However, long-term results are unknown and anatomical accommodation to fenestrated stent grafts with potential misalignments between the fenestrations and the target vessels because of sizing approximation or aneurysm shrinkage over time has not been previously described in literature. The aim of this study was to assess the short- and longterm positional change of renal arteries after FEVAR in the axial, sagittal, and coronal planes, and the consequences on renal stent patency as these changes may induce renal stent stenosis or occlusion. METHODS All fenestrated custom-made devices implanted at a single institution (Aortic Centre, University Hospital Centre of Lille, France) between October 2004 and January 2014 were identified from a prospectively maintained electronic endovascular aortic database. Patient inclusion Inclusion criteria for enrolment in the study were juxtarenal, pararenal, or thoraco-abdominal aortic aneurysms repaired with a custom-made device designed with fenestrations for at least one renal artery and the superior mesenteric artery (SMA) in a procedure performed more than 2 years ago; 1 mm thickness or less computed tomography angiography (CTA) of the chest, abdomen, and pelvis, available pre-operatively, post-operatively (within 6 months after the procedure), and during follow-up (at least 1 year after the procedure). Patients with branches or a scallop for the SMA or the renals were excluded, as well as patients with poor quality CTA, or either the post-operative (within 6 months) or the last CTA during follow-up (performed after 1 year) not available. Sizing and implantation of the stent grafts The sizing of the stent grafts was performed either by the manufacturer (Cook’s European planning centre, London, UK)
B. Maurel et al.
or by an experienced vascular surgeon who reviewed and approved the graft plan in every case (S.H.). The same experienced vascular surgeon performed or supervised all of the procedures (S.H.). Prior to December 2012, all cases were performed in an operating room with a mobile motorised CArm (OEC 9900 Elite MD; GE OEC Medical Systems, Inc., Salt Lake City, UT, USA). More recently, they have been performed in a dedicated hybrid operating room (Discovery; GE Healthcare, Chalfont St Giles, UK) under image fusion guidance. The endovascular devices were all custom-made and designed with fenestrations for the renal arteries and the SMA. A detailed description of the implantation procedure for complex endovascular aortic repair has been published previously.9 Imaging analysis For each patient, the high resolution pre-operative CTA dedicated to stent graft sizing, the post-operative CTA, and the last CTA available during follow-up, were analysed using the dedicated 3-dimensional workstation AquariusNET software (TeraRecon Inc., San Mateo, CA, USA). One trained vascular surgeon (L.Y.) has performed the measurements after the completion of an intra-operator concordance test (details in Statistical analysis). The following criteria were extracted from the CTA for analysis: left (LRA) and right renal artery (RRA) angulation; clock face position of the LRA and the RRA according to the SMA; and length between the SMA and the LRA and RRA (Fig. 1). The LRA, RRA, and SMA ostium were defined as the interface between the aorta and each target vessel. Segmentation of the aorta and the target vessels was performed as previously described,10 using a semi-automated centreline generated from the thoracic aorta to the aortic bifurcation. The centreline was assessed with multiplanar reconstruction views perpendicular to the centreline of flow, and then manually edited if necessary. Accessory renal arteries were not evaluated. Renal artery angulation The assessment of pre- and post-operative renal artery angulation was performed according to the technique described by Conway et al.11 The angle between the aortic centreline and the renal artery implantation was measured using the angular measurement tool provided by the workstation. A positive or negative renal artery implantation angle was defined as, respectively, above or below the horizontal plane perpendicular to the aortic centreline. This measurement was performed for both main renal arteries. Renal artery clock position Pre-operative renal artery orientation in a circumferential position was assessed using the graft plan with the SMA considered as the central position at 12:00 o’clock and the clock position of each renal artery amended according to the SMA. The post-operative measurements were performed on the CTA, using the aortic centreline reconstruction and the clock position tool, with 12 h gradations over 360 . The measurements were performed relative to the SMA position considered as the central position at 12:00 o’clock and to the nearest
Renal Changes after FEVAR
97
Figure 1. Imaging analysis. The following measurements were collected, using a semi-automated centreline generated from the thoracic aorta to the aortic bifurcation and manually edited if necessary: (A) right renal artery (RRA) angulation according to the centreline; (B) clock face position of the RRA after adjustment of the superior mesenteric artery (SMA) at 12 o’clock; (C) length between the SMA and the RRA.
15 min (equivalent to 7.5 ). Then clock positions were converted to degrees for analysis considering 12:00 o’clock as 0 . Renal artery distance to the SMA The pre-operative distance between renal arteries and SMA was collected from the graft plan using the distances from proximal edge. Post-operative measurements were performed using the stretch view generated by the centreline. The longitudinal vessel separation was manually measured using the centre of the SMA ostium as the reference point. The image was rotated on its centreline axis to identify the optimal view of the centre of each renal artery ostium.5 Statistical analysis Data and variations for each patient’s renal artery were reported as median and 25the75th interquartile ranges, and comparisons within each patient’s renal artery were performed using the non-parametric ManneWhitney test for non-normally distributed continuous variables. Renals angulations were presented in degrees ( ), renals clock positions in degrees ( ) and in hours:minutes (with 7.5 ¼ 00:15), and distance between renals and SMA in millimetres (mm). Variables (renals angulations, renals clock positions and distances relative to the SMA) were compared between the pre-operative and the post-operative CTA (performed within 6 months of the procedure), between the pre-operative and the last CTA during follow-up (performed at least 1 year after the procedure), and between the post-operative and the last CTA, for each patient’s renal arteries whenever complete dataset imaging was available.
The whole spectrum of differences between the absolute value (delta) of the variation between pre- and postoperative and follow-up CTA were illustrated using box plot figures. Intra-observer reliability was measured with the coefficient of variation (CV). Using a random sample of 10 patients, the observer evaluated each patient three times. The CV was calculated for both renal angles and all distances. A distribution with CV ¼ 0 is considered with no variance. A CV < 0.1 was required for the observer to begin with the study measurements. Values of p < .05 were considered to be statistically significant. All analyses were conducted using Epi Info 3.5.3. software (Centers for Disease Control and Prevention, Atlanta, GA, USA). RESULTS From October 2004 to January 2014, 100 patients met the inclusion criteria (Fig. 2). For all included patients the graft plan was available, and 86% of the CTA imaging was available for accurate analysis: 58 patients had both postoperative and last follow-up CTA; 21 patients had only post-operative CTA available (<6 months after surgery); and 21 patients had only last follow-up CTA available (>12 months after surgery). The median follow-up was 27.3 months (22.7e50.1). No renal occlusions occurred among these patients. The same type of bridging stent (Advanta V12, Atrium Medical Inc., Hudson, NH, USA) was used throughout the course of the study. Fig. 3 illustrates the observed changes in renal angulations and height for one
98
B. Maurel et al.
Figure 2. Flow chart of the patients included. SMA ¼ superior mesenteric artery; CTA ¼ computed tomography angiography.
patient over time. Among the 41 patients repaired with a 3fenestration stent graft but excluded from the analysis because of poor quality or missing imaging, five renal artery stents occluded and two required an additional procedure for in stent stenosis.
Coefficient of variation (CV) Among the 10 patients used to assess the intra-observer reliability, the median coefficient of variation was 0.08 for the renal angulation, 0 for the renal clock position, and 0.04 for the height between the renal arteries and the SMA.
Renal artery angulation The renal arteries assumed different angulations on preoperative compared with post-operative values, 28 (39 ; 13 ) and 11 (19 ; 0 ), respectively, on the left, p ¼ .0001, and 23 (42 ; 12 ) and 13 (24 ; 4 ), respectively, on the right, p ¼ .0001, but no significant change thereafter. These values reflect an upward positional change, and a delta (absolute value of change) between pre- and last follow-up CTA of 17 (12 ; 29 ). During the follow-up, 10 renal arteries (6.3%) presented a movement in trunk angulation greater than 45 (maximum 75 ). Details are available in Tables 1 and 2, and are illustrated in Fig. 4.
Figure 3. Observed changes over time for one patient. Pre-operative (left column), post-operative (intermediate column), and last followup (right column) CTA illustrating changes over time of the left renal artery angulation (first row) and clock face position (second row).
Renal Changes after FEVAR
99
Renal artery clock position
DISCUSSION
A significant positional change was found between the preand post-operative CTA (00:15 [00:00; 00:30], p < .05 on the left; 00:15 [00:15; 00:30], p < .001 on the right), without significant change thereafter (p ¼ .18 on the left and p ¼ .28 on the right). The movement between pre-operative and last follow-up CTA was 00:30 [00:15; 00:45] and 00:15 [00:15; 00:15] for the left and right renal arteries, respectively, with six renal arteries presenting a delta of 01:00 and two other arteries a delta of 01:15. A trend toward an anterior positional change was seen (from 02:30 to 02:15 on the left, p < .0001, and from 09:45 to 10:00 on the right, p < .0001). Details are available in Tables 1 and 2, and are illustrated in Fig. 5.
This study describes the change in position of the renal arteries after FEVAR with a 27.3 months [22.7; 50.1] follow-up. The present data reveal a tolerance of the renal arteries to permanent changes in angulation of approximately 17 upward trunk positional change and of 15e30 min anterior ostial positional change, mainly during the post-operative course, without adverse consequences on patency. The distance between the target vessels remains stable over time. These results suggest that with proper planning and a custom-made device implantation, there is a significant ostium and trunk positional change of both renal arteries after the implantation without adverse events. During the follow-up there is no significant change in position over time. Longitudinal distance between aortic side branches and circumferential orientation of the target vessel ostia on the orthogonal aortic cross-section, the “clock position”, are the two most important measurements required to design a fenestrated body, to allow target vessel perfusion while achieving aneurysm exclusion and long-term durability. The sizing and design of branched and fenestrated stent grafts is
Renal artery to SMA distance No significant change was found in the distance from the renal arteries to the SMA. Seven (4.5%) renal arteries presented 5 mm or more of movement to the SMA on the follow-up CTA (maximum 14 mm). Details are available in Tables 1 and 2, and are illustrated in Fig. 6.
Table 1. Renal arteries values in angulation, clock face position, and height over time, including 58 patients with complete imaging, and 42 patients with either post-operative (N ¼ 21) or follow-up (N ¼ 21) imaging. Pre-op CTA (n ¼ 79) Post-op CTA (n ¼ 79) p Pre-op CTA (n ¼ 79) Last CTA (n ¼ 79) p Renal angulation [ ] LRA 28 [39; 13] 11 [19; 0] <.0001 25.5 [37; 15] 11 [16; 0] <.0001 RRA 23 [42; 12] 13 [24; 4] .0001 27 [40; 12] 10 [23; 0] <.0001 Both 26 [41; 13] 11 [22; 2] <.0001 26 [38; 13] 10 [19; 0] <.0001 Renal clock face position [hour:minute] LRA 02:45 [02:15; 03:00] 02:30 [02:00; 03:00] .0369 02:30 [02:15; 03:00] 02:15 [02:00; 02:30] <.0001 RRA 09:45 [09:30; 10:00] 10:00 [09:30; 10:15] .0006 09:45 [09:30; 10:00] 10:00 [09:45; 10:15] <.0001 Renal height according to the SMA [mm] LRA 14 [9; 18] 13 [8; 18] .61 12 [8; 19] 12.6 [8; 18] .8143 RRA 11 [7; 16] 10 [6.6; 16] .65 10 [5; 15] 10 [5; 14.6] .7122 Both 12 [9; 17] 12 [7; 17] .52 11 [7; 16] 11 [6.2; 16] .6462 Results are expressed in median and 25the75th interquartile ranges. LRA ¼ left renal artery; RRA ¼ right renal artery; SMA ¼ superior mesenteric artery; both ¼ both renal arteries; pre-op ¼ pre-operative; post-op ¼ post-operative; CTA ¼ computed tomography angiography. Table 2. Absolute change (delta) over time in renal artery angulation, clock face position and height, among the 58 patients with complete imaging N ¼ 58 patients. Delta pre-op/post-op p Delta pre-op/last FU p Delta post-op/last FU p Renal angulation ( ) LRA 17 [6; 30] <.001 17 [12; 29] <.001 7 [3; 13] .58 RRA 18 [10; 29] <.001 18 [11.5; 28; 5] <.001 6.5 [2; 13] .7 Both renals 17 [7.5; 29] <.001 17 [12; 29] <.001 7 [2; 13] .5 Renal clock face position (hour:minute) LRA 00:15 [00:00; 00:30] .03 00:30 [00:15; 00:45] <.001 00:15 [00:00; 00:30] .18 RRA 00:15 [00:15; 00:30] <.001 00:15 [00:15; 00:15] <.001 00:15 [00:00; 00:15] .28 Both renals 00:15 [00:15; 00:30] <.001 00:15 [00:15; 00:30] <.001 00:15 [00:00; 00:15] .14 Renal height according to the SMA (mm) LRA 1.4 [1; 3] .54 1.5 [1; 2.4] .84 1 [0.2; 1.5] .69 RRA 1.55 [1; 3] .6 1.7 [1; 3] .70 0.95 [0.1; 1.7] .92 Both renals 1.5 [1; 3] .43 1.65 [1; 2.5] .63 1 [0.2; 1.5] .77 Results are expressed in median and 25the75th interquartile ranges. LRA ¼ left renal artery; RRA ¼ right renal artery; SMA ¼ superior mesenteric artery; both ¼ combination of measurements from both renal arteries; CTA ¼ computed tomography angiography; delta preop/post-op ¼ movement between the pre-operative and the post-operative; delta post-op/last CTA ¼ movement between the postoperative and the last CTA during follow-up; CTA; delta pre/last ¼ movement between the pre-operative and the last CTA; p ¼ analysis of the movement between the post-operative and the last follow-up CTA.
100
B. Maurel et al.
Figure 4. Box plot figure of the absolute value of change (delta) in renal artery angulation expressed in degree. PRE-OP/POSTOP ¼ absolute value of change between the pre-operative and the post-operative CTA; PRE-OP/FU ¼ absolute value of change between the pre-operative and the last CTA during follow-up; POSTOP/FU ¼ absolute value of change between the post-operative and the last CTA during follow-up.
Figure 6. Box plot figure of the absolute value of change (delta) in renal artery distance to the superior mesenteric artery (SMA) expressed in millimetres. PRE-OP/POST-OP ¼ absolute value of change between the pre-operative and the post-operative CTA; PREOP/FU ¼ absolute value of change between the pre-operative and the last CTA during follow-up; POST-OP/FU ¼ absolute value of change between the post-operative and the last CTA during follow-up.
mostly performed by the manufacturer in a planning centre, but should be always reviewed by the surgeon before manufacture. Other than the vessel anatomy, several parameters may impact the design of the device. Technically, to meet manufacturing constraints, a minimum separation is required between two fenestrations which depend on the graft diameter, and the fenestration has to be located between the stent struts to be “strut free”. Moreover, fenestrations are not located on clock position alone. Rather, they are located based on arc length determined from clock position and inner vessel diameter. This may not affect postoperative placement, but at the time of placement the
graft is compressed more on the posterior than the anterior side. As the graft expands over time, and largely on the posterior side, this would move fenestrations anteriorly. Additionally, when the device is unsheathed, single or double diameter reducing ties are linked to the trigger wires to partially deploy the graft and allow the operator minor adjustment of the height and rotation. During that time, the fenestrations are facing more posteriorly than when the device is fully deployed, and to facilitate the catheterisation of target vessels, the renal fenestrations may be designed slightly anteriorly. Finally, steep renal angulation may challenge both the accurate visualisation of the ostia and the sizing. All of these reasons, in combination with 15e20% of graft oversizing, contribute to make the sizing of a fenestrated stent graft partly intuitive. This may explain why significant intra- and inter-observer discrepancy have been reported. Oshin et al.5 reported an analysis of 25 FEVAR patients independently reviewed by two experienced observers, each using two different standardised protocols (Leonardo or Aquarius workstation) for each patient. Intra-observer variability was minimal, with insignificant changes regardless of sizing technique, but inter-observer variability was greater, with 18% more than 3 mm discrepancy in longitudinal vessel separation and 12% more than 30 min of discrepancy in clock measurements. In another study reported by Banno et al.,7 two independent experienced endovascular surgeons reviewed 268 pre-operative CTA and compared their measurements with the fenestrated graft plans performed by the manufacturer. A difference of more than 5 mm in visceral artery distance was noted in 16.4%, and of more than 45 min in vessel orientation in 3.6% for the RRA and 6.2% for the LRA. However, the correlation between various sizings and success rate or long-term target vessel patency is not known, and additionally the centrelines used for measurement of ostium location are likely different from the centreline use for graft design, using different adjustments.The
Figure 5. Box plot figure of the absolute value of change (delta) in renal artery clock position expressed in degrees. PRE-OP/POSTOP ¼ absolute value of change between the pre-operative and the post-operative CTA; PRE-OP/FU ¼ absolute value of change between the pre-operative and the last CTA during follow-up; POSTOP/FU ¼ absolute value of change between the post-operative and the last CTA during follow-up.
Renal Changes after FEVAR
present study is the first analysis reporting a long-term permanent anterior change in renal arteries clock position of 30 min (15e45) for the LRA and 15 min (15e15) for the RRA, without adverse renal events. Interestingly, most of the changes were noted in the post-operative course, thus probably related to sizing accommodation and to the manner in which grafts are manufactured and fenestrations located. These results suggest that an accommodation to sizing errors up to 30 min in clock position may be considered acceptable and is unlikely to lead to a clinically significant event. Recently there has been interest in developing “off the shelf” fenestrated stent grafts that would enable treatment of patients with juxtarenal aneurysms and favourable anatomy, reducing both the costs and manufacturing delays of a custom-made device.8,12 In an anatomical study, Sobocinski et al.12 reported that 70% of patients with a pararenal aneurysm may be suitable to be treated with one of the two configurations of the Cook Medical off the shelf stent graft (pbranch). However, some have argued that patients presenting with symptomatic aneurysms, who are more likely to benefit from this technology, usually have more complex anatomy that may not fit the requirement to have a successful repair with an off the shelf fenestrated stent graft.13 Early clinical experience using the off the shelf Cook p-Branch fenestrated device (Cook Medical, Bloomington, IN, USA) has been reported by Kitagawa et al.8 among 16 patients including two symptomatic aneurysms. Results were promising with 100% technical success, only one renal artery occlusion at 3 months, and an overall applicability of the p-branch over 70%. The position of the right renal artery was the primary reason for exclusion. The results of the present study suggest that sizing error accommodation for the renal clock position of up to 15e 30 min is common after custom-made device implantation without adverse consequences, and that the bridging stent induces a 17 trunk positional change. These results may be a step forward for use of off the shelf fenestrated device suggesting tolerance of the anatomy to sizing approximations, also considering that in the future, with the development of lower profile and more flexible bridging stents and stent grafts, the degree of flexibility tolerated by the anatomy is likely to increase without adverse effects. However, it could also be hypothesised that given these results showing significant anatomical change after use of custom-made devices designed to fit perfectly each patient’s anatomy, off the shelf devices may lead to even more post-operative positional modifications in an unpredictable way and worse results than expected regarding organ ischaemia and infarction. This study has several limitations. First, the patient images were analysed retrospectively, and patients without imaging available, poor quality or non-contrast CTA were excluded from the study.Thus, this analysis is not exhaustive as patients with renal impairment or adverse renal events, including five patients with renal stent occlusion, were excluded from the analysis. However, the present goal was to assess long-term renal fenestration movement among patients with renal artery patency, to estimate how much inaccuracy in sizing can occur without adverse consequences. Second, to increase the number of patients, all patients with either a CTA in the post-
101
operative course and/or during the follow-up were included, even if only 58% of the patients had both post-operative and long-term high quality imaging available. Consequently, the post-operative group is not identical to the follow-up group, even if of the same size. Therefore, only the 58 patients with all imaging available were used to assess and compare the movements among the pre-operative, post-operative, and last follow-up CTA. Finally, only one experienced observer performed all of the measurements, thus these results do not reflect potential intra-observer variability. CONCLUSION The present study demonstrates a tolerance of the anatomy to fenestrated device sizing inaccuracy, as well as to anatomical change over time on the left renal artery. These results may have implications for “off the shelf” devices in favourable anatomies that could have a major role to play in acute complex aneurysm repair, as well as in elective cases to reduce cost and standardise the procedure. Future research in this area should now include a cohort of patients with adverse renal events after FEVAR to determine if risk factors for renal fenestration occlusion can be determined, as lessons should be learned to overcome this issue and potentially improve practice. CONFLICT OF INTEREST A.H. is a consultant for GE Healthcare; T.M.M. is a consultant for Cook Medical; S.H. is a consultant for Cook Medical and GE Healthcare. FUNDING This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. REFERENCES 1 Carpenter JP, Baum RA, Barker CF, Golden MA, Mitchell ME, Velazquez OC, et al. Impact of exclusion criteria on patient selection for endovascular abdominal aortic aneurysm repair. J Vasc Surg 2001;34(6):1050e4. 2 Greenberg RK, Sternbergh 3rd WC, Makaroun M, Ohki T, Chuter T, Bharadwaj P, et al. Fenestrated Investigators. Intermediate results of a United States multicenter trial of fenestrated endograft repair for juxtarenal abdominal aortic aneurysms. J Vasc Surg 2009;50(4):730e7. 3 Katsargyris A, Verhoeven EL. Endovascular strategies for infrarenal aneurysms with short necks. J Cardiovasc Surg (Torino) 2013;54(1 Suppl. 1):21e6. 4 Hertault A, Haulon S, Lee JT. Debate: whether branched/ fenestrated endovascular aneurysm repair procedures are better than snorkels, chimneys, or periscopes in the treatment of most thoracoabdominal and juxtarenal aneurysms. J Vasc Surg 2015;62(5):1357e65. 5 Oshin OA, England A, McWilliams RG, Brennan JA, Fisher RK, Vallabhaneni SR. Intra- and interobserver variability of target vessel measurement for fenestrated endovascular aneurysm repair. J Endovasc Ther 2010;17(3):402e7. 6 Malkawi AH, Resch TA, Bown MJ, Manning BJ, Poloniecki JD, Nordon IM, et al. Sizing fenestrated aortic stent-grafts. Eur J Vasc Endovasc Surg 2011;41(3):311e6.
102 7 Banno H, Kobeiter H, Brossier J, Marzelle J, Presles E, Becquemin JP. Inter-observer variability in sizing fenestrated and/or branched aortic stent-grafts. Eur J Vasc Endovasc Surg 2014;47(1):45e52. 8 Kitagawa A, Greenberg RK, Eagleton MJ, Mastracci TM. Zenith pbranch standard fenestrated endovascular graft for juxtarenal abdominal aortic aneurysms. J Vasc Surg 2013;58(2):291e300. 9 Guillou M, Bianchini A, Sobocinski J, Maurel B, D’elia P, Tyrrell M, et al. Endovascular treatment of thoracoabdominal aortic aneurysms. J Vasc Surg 2012;56(1):65e73. 10 Goel VR, Greenberg RK, Greenberg DP. Mathematical analysis of DICOM CT datasets: can endograft sizing be automated for complex anatomy? J Vasc Surg 2008;47(6):1306e12.
M. Elkawafi and U.J. Kirkpatrick 11 Conway BD, Greenberg RK, Mastracci TM, Hernandez AV, Coscas R. Renal artery implantation angles in thoracoabdominal aneurysms and their implications in the era of branched endografts. J Endovasc Ther 2010;17(3):380e7. 12 Sobocinski J, d’Utra G, O’Brien N, Midulla M, Maurel B, Guillou M, et al. Off-the-shelf fenestrated endografts: a realistic option for more than 70% of patients with juxtarenal aneurysms. J Endovasc Ther 2012;19(2):165e72. 13 Nordon IM, Hinchliffe RJ, Manning B, Ivancev K, Holt PJ, Loftus IM, et al. Toward an “off-the-shelf” fenestrated endograft for management of short-necked abdominal aortic aneurysms: an analysis of current graft morphological diversity. J Endovasc Ther 2010;17(1):78e85.
Eur J Vasc Endovasc Surg (2017) 53, 102
COUP D’OEIL
Sacotomy if All Else Fails? M. Elkawafi *, U.J. Kirkpatrick Wrexham Maelor Hospital, Wrexham, UK
A 73-year-old male underwent endovascular repair of a 6-cm infrarenal aortic aneurysm using a Medtronic Endurant device (Medtronic, Wrexham, Wales, UK). Surveillance computed tomography scans demonstrated type 2 endoleak from a lumbar artery and the sac size increased by >1 cm. Subsequent attempts at lumbar artery branch embolisation were unsuccessful and percutaneous transabdominal glue injections were ineffective. The aneurysm enlarged to 7.7 cm at 27 months. The patient then developed episodes of abdominal pain; transperitoneal mini-laparotomy, sacotomy, and oversew of the lumbar artery were performed, leaving the stent graft undisturbed. After surgery, the sac diameter measured 5.3 cm. The patient remains well 5 years after the sacotomy procedure. * Corresponding author. E-mail address: Mohamed.elkawafi@hotmail.com (M. Elkawafi). 1078-5884/Ó 2016 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejvs.2016.10.027
Changes in Renal Anatomy After Fenestrated Endovascular Aneurysm Repair5 B. Maurel a,b, Y. Lounes a, M. Amako a, D. Fabre a, A. Hertault a, J. Sobocinski a, R. Spear a, R. Azzaoui a, T.M. Mastracci b, S. Haulon a,* a b
Aortic Center, Hôpital Cardiologique, CHRU de Lille, France Aortic Team, Royal Free London NHS Foundation Trust, London, UK
WHAT THIS PAPER ADDS Fenestrated devices have become the standard of care for juxtarenal aneurysms. Knowing the impact that variability in renal stent position has on long-term outcomes may influence clinical pathways. This study explores the changing positions of renal anatomy after fenestrated repair.
Objective: To assess short- and long-term movement of renal arteries after fenestrated endovascular aortic repair (FEVAR). Methods: Consecutive patients who underwent FEVAR at one institution with a custom-made device designed with fenestrations for the superior mesenteric (SMA) and renal arteries, a millimetric computed tomography angiography (CTA), and a minimum of 2 years’ follow-up were included. Angulation between renal artery trunk and aorta, clock position of the origin of the renal arteries, distance between renal arteries and SMA, and target vessel occlusion were retrospectively collected and compared between the pre-operative, post-operative (<6 months), and last (>12 months) CTA. Results: From October 2004 to January 2014, 100 patients met the inclusion criteria and 86% of imaging was available for accurate analysis. Median follow-up was 27.3 months (22.7e50.1). There were no renal occlusions. A significant change was found in the value of renal trunk angulation of both renal arteries on post-operative compared with pre-operative CTA (17 difference upward [7.5e29], p < .001), but no significant change thereafter (p ¼ .5). Regarding renal clock positions (7.5 of change equivalent to 15 min of renal ostial movement): significant anterior change was found between post-operative and pre-operative CTA (15 min [0e 30], p ¼ .03 on the left and 15 min [15e30], p < .001 on the right), without significant change thereafter (15 min [0e30], p ¼ .18 on the left and 15 min [0e15] on the right, p ¼ .28). No changes were noted on the distance between renal and SMA ostia (difference of 1.65 mm [1e2.5], p ¼ .63). Conclusion: The renal arteries demonstrate tolerance to permanent changes in angulation after FEVAR of approximately 17 upward trunk movement and of 15e30 min ostial movement without adverse consequences on patency after a median of more than 2 years’ follow-up. The distance between the target vessels remained stable over time. These results may suggest accommodation to sizing errors and thus a compliance with off the shelf devices in favourable anatomies. Ó 2016 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. Article history: Received 18 July 2016, Accepted 24 October 2016, Available online 24 November 2016 Keywords: FEVAR, Renal angulation, Renal clock position, Sizing, Off the shelf devices
INTRODUCTION Over the past two decades, endovascular repair of abdominal aortic aneurysms (EVAR) has gained popularity as a treatment option that offers many advantages over 5 This study has been presented in the plenary session at the ESVS 30th Annual Meeting in Copenhagen, Denmark, on 28e30 September 2016 (abstract n 1357). * Corresponding author. Aortic Centre, CHRU de Lille, INSERM U1008, Université Lille Nord de France, 59037 Lille Cedex, France. E-mail address: [email protected] (S. Haulon). 1078-5884/Ó 2016 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejvs.2016.10.019
open surgical repair. However, specified anatomical characteristics are required to achieve durable outcomes, and up to 40% of patients are unsuitable for traditional infrarenal EVAR, mainly because of a hostile proximal neck anatomy.1,2 To overcome these limitations, use of a fenestrated stent graft for fenestrated endovascular aneurysm repair (FEVAR) currently represents the most validated and reliable endovascular option to treat short neck or pararenal aneurysms, with high technical success rates (99%), low operative mortality (3.5%), and excellent mid- and long-term target vessel patency (97%).3,4 Fenestrated stent grafts are custom-made to fit each patient’s anatomy, implying a delay for planning,
96
production, and implantation from 6 to 8 weeks. The success of the procedure relies on appropriate sizing of the stentgraft, including accurate determination of target vessel clock face position, distance from proximal stentgraft edge, and aortic diameter, to ensure aneurysm exclusion and target vessel patency. Sizing is commonly performed by the manufacturer in a centralised planning centre. If experienced surgeons do their own planning, it is a complex process that includes learning anatomical limitations to meet manufacturing constraints and requires a 3-dimensional workstation. Moreover, concerns have been raised over intra- and inter-observer variability and potential consequences.5e7 As not all patients are able to wait for a customised graft, off the shelf fenestrated devices, including three fenestrations for the superior mesenteric artery (SMA) and renal arteries, and a scallop for the coeliac trunk (CT), available for immediate use and designed to accommodate the majority of the patients are under investigation with encouraging early results.8 It has been shown to be applicable in 70% of the patients, with the right renal artery as a primary cause of exclusion. However, long-term results are unknown and anatomical accommodation to fenestrated stent grafts with potential misalignments between the fenestrations and the target vessels because of sizing approximation or aneurysm shrinkage over time has not been previously described in literature. The aim of this study was to assess the short- and longterm positional change of renal arteries after FEVAR in the axial, sagittal, and coronal planes, and the consequences on renal stent patency as these changes may induce renal stent stenosis or occlusion. METHODS All fenestrated custom-made devices implanted at a single institution (Aortic Centre, University Hospital Centre of Lille, France) between October 2004 and January 2014 were identified from a prospectively maintained electronic endovascular aortic database. Patient inclusion Inclusion criteria for enrolment in the study were juxtarenal, pararenal, or thoraco-abdominal aortic aneurysms repaired with a custom-made device designed with fenestrations for at least one renal artery and the superior mesenteric artery (SMA) in a procedure performed more than 2 years ago; 1 mm thickness or less computed tomography angiography (CTA) of the chest, abdomen, and pelvis, available pre-operatively, post-operatively (within 6 months after the procedure), and during follow-up (at least 1 year after the procedure). Patients with branches or a scallop for the SMA or the renals were excluded, as well as patients with poor quality CTA, or either the post-operative (within 6 months) or the last CTA during follow-up (performed after 1 year) not available. Sizing and implantation of the stent grafts The sizing of the stent grafts was performed either by the manufacturer (Cook’s European planning centre, London, UK)
B. Maurel et al.
or by an experienced vascular surgeon who reviewed and approved the graft plan in every case (S.H.). The same experienced vascular surgeon performed or supervised all of the procedures (S.H.). Prior to December 2012, all cases were performed in an operating room with a mobile motorised CArm (OEC 9900 Elite MD; GE OEC Medical Systems, Inc., Salt Lake City, UT, USA). More recently, they have been performed in a dedicated hybrid operating room (Discovery; GE Healthcare, Chalfont St Giles, UK) under image fusion guidance. The endovascular devices were all custom-made and designed with fenestrations for the renal arteries and the SMA. A detailed description of the implantation procedure for complex endovascular aortic repair has been published previously.9 Imaging analysis For each patient, the high resolution pre-operative CTA dedicated to stent graft sizing, the post-operative CTA, and the last CTA available during follow-up, were analysed using the dedicated 3-dimensional workstation AquariusNET software (TeraRecon Inc., San Mateo, CA, USA). One trained vascular surgeon (L.Y.) has performed the measurements after the completion of an intra-operator concordance test (details in Statistical analysis). The following criteria were extracted from the CTA for analysis: left (LRA) and right renal artery (RRA) angulation; clock face position of the LRA and the RRA according to the SMA; and length between the SMA and the LRA and RRA (Fig. 1). The LRA, RRA, and SMA ostium were defined as the interface between the aorta and each target vessel. Segmentation of the aorta and the target vessels was performed as previously described,10 using a semi-automated centreline generated from the thoracic aorta to the aortic bifurcation. The centreline was assessed with multiplanar reconstruction views perpendicular to the centreline of flow, and then manually edited if necessary. Accessory renal arteries were not evaluated. Renal artery angulation The assessment of pre- and post-operative renal artery angulation was performed according to the technique described by Conway et al.11 The angle between the aortic centreline and the renal artery implantation was measured using the angular measurement tool provided by the workstation. A positive or negative renal artery implantation angle was defined as, respectively, above or below the horizontal plane perpendicular to the aortic centreline. This measurement was performed for both main renal arteries. Renal artery clock position Pre-operative renal artery orientation in a circumferential position was assessed using the graft plan with the SMA considered as the central position at 12:00 o’clock and the clock position of each renal artery amended according to the SMA. The post-operative measurements were performed on the CTA, using the aortic centreline reconstruction and the clock position tool, with 12 h gradations over 360 . The measurements were performed relative to the SMA position considered as the central position at 12:00 o’clock and to the nearest
Renal Changes after FEVAR
97
Figure 1. Imaging analysis. The following measurements were collected, using a semi-automated centreline generated from the thoracic aorta to the aortic bifurcation and manually edited if necessary: (A) right renal artery (RRA) angulation according to the centreline; (B) clock face position of the RRA after adjustment of the superior mesenteric artery (SMA) at 12 o’clock; (C) length between the SMA and the RRA.
15 min (equivalent to 7.5 ). Then clock positions were converted to degrees for analysis considering 12:00 o’clock as 0 . Renal artery distance to the SMA The pre-operative distance between renal arteries and SMA was collected from the graft plan using the distances from proximal edge. Post-operative measurements were performed using the stretch view generated by the centreline. The longitudinal vessel separation was manually measured using the centre of the SMA ostium as the reference point. The image was rotated on its centreline axis to identify the optimal view of the centre of each renal artery ostium.5 Statistical analysis Data and variations for each patient’s renal artery were reported as median and 25the75th interquartile ranges, and comparisons within each patient’s renal artery were performed using the non-parametric ManneWhitney test for non-normally distributed continuous variables. Renals angulations were presented in degrees ( ), renals clock positions in degrees ( ) and in hours:minutes (with 7.5 ¼ 00:15), and distance between renals and SMA in millimetres (mm). Variables (renals angulations, renals clock positions and distances relative to the SMA) were compared between the pre-operative and the post-operative CTA (performed within 6 months of the procedure), between the pre-operative and the last CTA during follow-up (performed at least 1 year after the procedure), and between the post-operative and the last CTA, for each patient’s renal arteries whenever complete dataset imaging was available.
The whole spectrum of differences between the absolute value (delta) of the variation between pre- and postoperative and follow-up CTA were illustrated using box plot figures. Intra-observer reliability was measured with the coefficient of variation (CV). Using a random sample of 10 patients, the observer evaluated each patient three times. The CV was calculated for both renal angles and all distances. A distribution with CV ¼ 0 is considered with no variance. A CV < 0.1 was required for the observer to begin with the study measurements. Values of p < .05 were considered to be statistically significant. All analyses were conducted using Epi Info 3.5.3. software (Centers for Disease Control and Prevention, Atlanta, GA, USA). RESULTS From October 2004 to January 2014, 100 patients met the inclusion criteria (Fig. 2). For all included patients the graft plan was available, and 86% of the CTA imaging was available for accurate analysis: 58 patients had both postoperative and last follow-up CTA; 21 patients had only post-operative CTA available (<6 months after surgery); and 21 patients had only last follow-up CTA available (>12 months after surgery). The median follow-up was 27.3 months (22.7e50.1). No renal occlusions occurred among these patients. The same type of bridging stent (Advanta V12, Atrium Medical Inc., Hudson, NH, USA) was used throughout the course of the study. Fig. 3 illustrates the observed changes in renal angulations and height for one
98
B. Maurel et al.
Figure 2. Flow chart of the patients included. SMA ¼ superior mesenteric artery; CTA ¼ computed tomography angiography.
patient over time. Among the 41 patients repaired with a 3fenestration stent graft but excluded from the analysis because of poor quality or missing imaging, five renal artery stents occluded and two required an additional procedure for in stent stenosis.
Coefficient of variation (CV) Among the 10 patients used to assess the intra-observer reliability, the median coefficient of variation was 0.08 for the renal angulation, 0 for the renal clock position, and 0.04 for the height between the renal arteries and the SMA.
Renal artery angulation The renal arteries assumed different angulations on preoperative compared with post-operative values, 28 (39 ; 13 ) and 11 (19 ; 0 ), respectively, on the left, p ¼ .0001, and 23 (42 ; 12 ) and 13 (24 ; 4 ), respectively, on the right, p ¼ .0001, but no significant change thereafter. These values reflect an upward positional change, and a delta (absolute value of change) between pre- and last follow-up CTA of 17 (12 ; 29 ). During the follow-up, 10 renal arteries (6.3%) presented a movement in trunk angulation greater than 45 (maximum 75 ). Details are available in Tables 1 and 2, and are illustrated in Fig. 4.
Figure 3. Observed changes over time for one patient. Pre-operative (left column), post-operative (intermediate column), and last followup (right column) CTA illustrating changes over time of the left renal artery angulation (first row) and clock face position (second row).
Renal Changes after FEVAR
99
Renal artery clock position
DISCUSSION
A significant positional change was found between the preand post-operative CTA (00:15 [00:00; 00:30], p < .05 on the left; 00:15 [00:15; 00:30], p < .001 on the right), without significant change thereafter (p ¼ .18 on the left and p ¼ .28 on the right). The movement between pre-operative and last follow-up CTA was 00:30 [00:15; 00:45] and 00:15 [00:15; 00:15] for the left and right renal arteries, respectively, with six renal arteries presenting a delta of 01:00 and two other arteries a delta of 01:15. A trend toward an anterior positional change was seen (from 02:30 to 02:15 on the left, p < .0001, and from 09:45 to 10:00 on the right, p < .0001). Details are available in Tables 1 and 2, and are illustrated in Fig. 5.
This study describes the change in position of the renal arteries after FEVAR with a 27.3 months [22.7; 50.1] follow-up. The present data reveal a tolerance of the renal arteries to permanent changes in angulation of approximately 17 upward trunk positional change and of 15e30 min anterior ostial positional change, mainly during the post-operative course, without adverse consequences on patency. The distance between the target vessels remains stable over time. These results suggest that with proper planning and a custom-made device implantation, there is a significant ostium and trunk positional change of both renal arteries after the implantation without adverse events. During the follow-up there is no significant change in position over time. Longitudinal distance between aortic side branches and circumferential orientation of the target vessel ostia on the orthogonal aortic cross-section, the “clock position”, are the two most important measurements required to design a fenestrated body, to allow target vessel perfusion while achieving aneurysm exclusion and long-term durability. The sizing and design of branched and fenestrated stent grafts is
Renal artery to SMA distance No significant change was found in the distance from the renal arteries to the SMA. Seven (4.5%) renal arteries presented 5 mm or more of movement to the SMA on the follow-up CTA (maximum 14 mm). Details are available in Tables 1 and 2, and are illustrated in Fig. 6.
Table 1. Renal arteries values in angulation, clock face position, and height over time, including 58 patients with complete imaging, and 42 patients with either post-operative (N ¼ 21) or follow-up (N ¼ 21) imaging. Pre-op CTA (n ¼ 79) Post-op CTA (n ¼ 79) p Pre-op CTA (n ¼ 79) Last CTA (n ¼ 79) p Renal angulation [ ] LRA 28 [39; 13] 11 [19; 0] <.0001 25.5 [37; 15] 11 [16; 0] <.0001 RRA 23 [42; 12] 13 [24; 4] .0001 27 [40; 12] 10 [23; 0] <.0001 Both 26 [41; 13] 11 [22; 2] <.0001 26 [38; 13] 10 [19; 0] <.0001 Renal clock face position [hour:minute] LRA 02:45 [02:15; 03:00] 02:30 [02:00; 03:00] .0369 02:30 [02:15; 03:00] 02:15 [02:00; 02:30] <.0001 RRA 09:45 [09:30; 10:00] 10:00 [09:30; 10:15] .0006 09:45 [09:30; 10:00] 10:00 [09:45; 10:15] <.0001 Renal height according to the SMA [mm] LRA 14 [9; 18] 13 [8; 18] .61 12 [8; 19] 12.6 [8; 18] .8143 RRA 11 [7; 16] 10 [6.6; 16] .65 10 [5; 15] 10 [5; 14.6] .7122 Both 12 [9; 17] 12 [7; 17] .52 11 [7; 16] 11 [6.2; 16] .6462 Results are expressed in median and 25the75th interquartile ranges. LRA ¼ left renal artery; RRA ¼ right renal artery; SMA ¼ superior mesenteric artery; both ¼ both renal arteries; pre-op ¼ pre-operative; post-op ¼ post-operative; CTA ¼ computed tomography angiography. Table 2. Absolute change (delta) over time in renal artery angulation, clock face position and height, among the 58 patients with complete imaging N ¼ 58 patients. Delta pre-op/post-op p Delta pre-op/last FU p Delta post-op/last FU p Renal angulation ( ) LRA 17 [6; 30] <.001 17 [12; 29] <.001 7 [3; 13] .58 RRA 18 [10; 29] <.001 18 [11.5; 28; 5] <.001 6.5 [2; 13] .7 Both renals 17 [7.5; 29] <.001 17 [12; 29] <.001 7 [2; 13] .5 Renal clock face position (hour:minute) LRA 00:15 [00:00; 00:30] .03 00:30 [00:15; 00:45] <.001 00:15 [00:00; 00:30] .18 RRA 00:15 [00:15; 00:30] <.001 00:15 [00:15; 00:15] <.001 00:15 [00:00; 00:15] .28 Both renals 00:15 [00:15; 00:30] <.001 00:15 [00:15; 00:30] <.001 00:15 [00:00; 00:15] .14 Renal height according to the SMA (mm) LRA 1.4 [1; 3] .54 1.5 [1; 2.4] .84 1 [0.2; 1.5] .69 RRA 1.55 [1; 3] .6 1.7 [1; 3] .70 0.95 [0.1; 1.7] .92 Both renals 1.5 [1; 3] .43 1.65 [1; 2.5] .63 1 [0.2; 1.5] .77 Results are expressed in median and 25the75th interquartile ranges. LRA ¼ left renal artery; RRA ¼ right renal artery; SMA ¼ superior mesenteric artery; both ¼ combination of measurements from both renal arteries; CTA ¼ computed tomography angiography; delta preop/post-op ¼ movement between the pre-operative and the post-operative; delta post-op/last CTA ¼ movement between the postoperative and the last CTA during follow-up; CTA; delta pre/last ¼ movement between the pre-operative and the last CTA; p ¼ analysis of the movement between the post-operative and the last follow-up CTA.
100
B. Maurel et al.
Figure 4. Box plot figure of the absolute value of change (delta) in renal artery angulation expressed in degree. PRE-OP/POSTOP ¼ absolute value of change between the pre-operative and the post-operative CTA; PRE-OP/FU ¼ absolute value of change between the pre-operative and the last CTA during follow-up; POSTOP/FU ¼ absolute value of change between the post-operative and the last CTA during follow-up.
Figure 6. Box plot figure of the absolute value of change (delta) in renal artery distance to the superior mesenteric artery (SMA) expressed in millimetres. PRE-OP/POST-OP ¼ absolute value of change between the pre-operative and the post-operative CTA; PREOP/FU ¼ absolute value of change between the pre-operative and the last CTA during follow-up; POST-OP/FU ¼ absolute value of change between the post-operative and the last CTA during follow-up.
mostly performed by the manufacturer in a planning centre, but should be always reviewed by the surgeon before manufacture. Other than the vessel anatomy, several parameters may impact the design of the device. Technically, to meet manufacturing constraints, a minimum separation is required between two fenestrations which depend on the graft diameter, and the fenestration has to be located between the stent struts to be “strut free”. Moreover, fenestrations are not located on clock position alone. Rather, they are located based on arc length determined from clock position and inner vessel diameter. This may not affect postoperative placement, but at the time of placement the
graft is compressed more on the posterior than the anterior side. As the graft expands over time, and largely on the posterior side, this would move fenestrations anteriorly. Additionally, when the device is unsheathed, single or double diameter reducing ties are linked to the trigger wires to partially deploy the graft and allow the operator minor adjustment of the height and rotation. During that time, the fenestrations are facing more posteriorly than when the device is fully deployed, and to facilitate the catheterisation of target vessels, the renal fenestrations may be designed slightly anteriorly. Finally, steep renal angulation may challenge both the accurate visualisation of the ostia and the sizing. All of these reasons, in combination with 15e20% of graft oversizing, contribute to make the sizing of a fenestrated stent graft partly intuitive. This may explain why significant intra- and inter-observer discrepancy have been reported. Oshin et al.5 reported an analysis of 25 FEVAR patients independently reviewed by two experienced observers, each using two different standardised protocols (Leonardo or Aquarius workstation) for each patient. Intra-observer variability was minimal, with insignificant changes regardless of sizing technique, but inter-observer variability was greater, with 18% more than 3 mm discrepancy in longitudinal vessel separation and 12% more than 30 min of discrepancy in clock measurements. In another study reported by Banno et al.,7 two independent experienced endovascular surgeons reviewed 268 pre-operative CTA and compared their measurements with the fenestrated graft plans performed by the manufacturer. A difference of more than 5 mm in visceral artery distance was noted in 16.4%, and of more than 45 min in vessel orientation in 3.6% for the RRA and 6.2% for the LRA. However, the correlation between various sizings and success rate or long-term target vessel patency is not known, and additionally the centrelines used for measurement of ostium location are likely different from the centreline use for graft design, using different adjustments.The
Figure 5. Box plot figure of the absolute value of change (delta) in renal artery clock position expressed in degrees. PRE-OP/POSTOP ¼ absolute value of change between the pre-operative and the post-operative CTA; PRE-OP/FU ¼ absolute value of change between the pre-operative and the last CTA during follow-up; POSTOP/FU ¼ absolute value of change between the post-operative and the last CTA during follow-up.
Renal Changes after FEVAR
present study is the first analysis reporting a long-term permanent anterior change in renal arteries clock position of 30 min (15e45) for the LRA and 15 min (15e15) for the RRA, without adverse renal events. Interestingly, most of the changes were noted in the post-operative course, thus probably related to sizing accommodation and to the manner in which grafts are manufactured and fenestrations located. These results suggest that an accommodation to sizing errors up to 30 min in clock position may be considered acceptable and is unlikely to lead to a clinically significant event. Recently there has been interest in developing “off the shelf” fenestrated stent grafts that would enable treatment of patients with juxtarenal aneurysms and favourable anatomy, reducing both the costs and manufacturing delays of a custom-made device.8,12 In an anatomical study, Sobocinski et al.12 reported that 70% of patients with a pararenal aneurysm may be suitable to be treated with one of the two configurations of the Cook Medical off the shelf stent graft (pbranch). However, some have argued that patients presenting with symptomatic aneurysms, who are more likely to benefit from this technology, usually have more complex anatomy that may not fit the requirement to have a successful repair with an off the shelf fenestrated stent graft.13 Early clinical experience using the off the shelf Cook p-Branch fenestrated device (Cook Medical, Bloomington, IN, USA) has been reported by Kitagawa et al.8 among 16 patients including two symptomatic aneurysms. Results were promising with 100% technical success, only one renal artery occlusion at 3 months, and an overall applicability of the p-branch over 70%. The position of the right renal artery was the primary reason for exclusion. The results of the present study suggest that sizing error accommodation for the renal clock position of up to 15e 30 min is common after custom-made device implantation without adverse consequences, and that the bridging stent induces a 17 trunk positional change. These results may be a step forward for use of off the shelf fenestrated device suggesting tolerance of the anatomy to sizing approximations, also considering that in the future, with the development of lower profile and more flexible bridging stents and stent grafts, the degree of flexibility tolerated by the anatomy is likely to increase without adverse effects. However, it could also be hypothesised that given these results showing significant anatomical change after use of custom-made devices designed to fit perfectly each patient’s anatomy, off the shelf devices may lead to even more post-operative positional modifications in an unpredictable way and worse results than expected regarding organ ischaemia and infarction. This study has several limitations. First, the patient images were analysed retrospectively, and patients without imaging available, poor quality or non-contrast CTA were excluded from the study.Thus, this analysis is not exhaustive as patients with renal impairment or adverse renal events, including five patients with renal stent occlusion, were excluded from the analysis. However, the present goal was to assess long-term renal fenestration movement among patients with renal artery patency, to estimate how much inaccuracy in sizing can occur without adverse consequences. Second, to increase the number of patients, all patients with either a CTA in the post-
101
operative course and/or during the follow-up were included, even if only 58% of the patients had both post-operative and long-term high quality imaging available. Consequently, the post-operative group is not identical to the follow-up group, even if of the same size. Therefore, only the 58 patients with all imaging available were used to assess and compare the movements among the pre-operative, post-operative, and last follow-up CTA. Finally, only one experienced observer performed all of the measurements, thus these results do not reflect potential intra-observer variability. CONCLUSION The present study demonstrates a tolerance of the anatomy to fenestrated device sizing inaccuracy, as well as to anatomical change over time on the left renal artery. These results may have implications for “off the shelf” devices in favourable anatomies that could have a major role to play in acute complex aneurysm repair, as well as in elective cases to reduce cost and standardise the procedure. Future research in this area should now include a cohort of patients with adverse renal events after FEVAR to determine if risk factors for renal fenestration occlusion can be determined, as lessons should be learned to overcome this issue and potentially improve practice. CONFLICT OF INTEREST A.H. is a consultant for GE Healthcare; T.M.M. is a consultant for Cook Medical; S.H. is a consultant for Cook Medical and GE Healthcare. FUNDING This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. REFERENCES 1 Carpenter JP, Baum RA, Barker CF, Golden MA, Mitchell ME, Velazquez OC, et al. Impact of exclusion criteria on patient selection for endovascular abdominal aortic aneurysm repair. J Vasc Surg 2001;34(6):1050e4. 2 Greenberg RK, Sternbergh 3rd WC, Makaroun M, Ohki T, Chuter T, Bharadwaj P, et al. Fenestrated Investigators. Intermediate results of a United States multicenter trial of fenestrated endograft repair for juxtarenal abdominal aortic aneurysms. J Vasc Surg 2009;50(4):730e7. 3 Katsargyris A, Verhoeven EL. Endovascular strategies for infrarenal aneurysms with short necks. J Cardiovasc Surg (Torino) 2013;54(1 Suppl. 1):21e6. 4 Hertault A, Haulon S, Lee JT. Debate: whether branched/ fenestrated endovascular aneurysm repair procedures are better than snorkels, chimneys, or periscopes in the treatment of most thoracoabdominal and juxtarenal aneurysms. J Vasc Surg 2015;62(5):1357e65. 5 Oshin OA, England A, McWilliams RG, Brennan JA, Fisher RK, Vallabhaneni SR. Intra- and interobserver variability of target vessel measurement for fenestrated endovascular aneurysm repair. J Endovasc Ther 2010;17(3):402e7. 6 Malkawi AH, Resch TA, Bown MJ, Manning BJ, Poloniecki JD, Nordon IM, et al. Sizing fenestrated aortic stent-grafts. Eur J Vasc Endovasc Surg 2011;41(3):311e6.
102 7 Banno H, Kobeiter H, Brossier J, Marzelle J, Presles E, Becquemin JP. Inter-observer variability in sizing fenestrated and/or branched aortic stent-grafts. Eur J Vasc Endovasc Surg 2014;47(1):45e52. 8 Kitagawa A, Greenberg RK, Eagleton MJ, Mastracci TM. Zenith pbranch standard fenestrated endovascular graft for juxtarenal abdominal aortic aneurysms. J Vasc Surg 2013;58(2):291e300. 9 Guillou M, Bianchini A, Sobocinski J, Maurel B, D’elia P, Tyrrell M, et al. Endovascular treatment of thoracoabdominal aortic aneurysms. J Vasc Surg 2012;56(1):65e73. 10 Goel VR, Greenberg RK, Greenberg DP. Mathematical analysis of DICOM CT datasets: can endograft sizing be automated for complex anatomy? J Vasc Surg 2008;47(6):1306e12.
M. Elkawafi and U.J. Kirkpatrick 11 Conway BD, Greenberg RK, Mastracci TM, Hernandez AV, Coscas R. Renal artery implantation angles in thoracoabdominal aneurysms and their implications in the era of branched endografts. J Endovasc Ther 2010;17(3):380e7. 12 Sobocinski J, d’Utra G, O’Brien N, Midulla M, Maurel B, Guillou M, et al. Off-the-shelf fenestrated endografts: a realistic option for more than 70% of patients with juxtarenal aneurysms. J Endovasc Ther 2012;19(2):165e72. 13 Nordon IM, Hinchliffe RJ, Manning B, Ivancev K, Holt PJ, Loftus IM, et al. Toward an “off-the-shelf” fenestrated endograft for management of short-necked abdominal aortic aneurysms: an analysis of current graft morphological diversity. J Endovasc Ther 2010;17(1):78e85.
Eur J Vasc Endovasc Surg (2017) 53, 102
COUP D’OEIL
Sacotomy if All Else Fails? M. Elkawafi *, U.J. Kirkpatrick Wrexham Maelor Hospital, Wrexham, UK
A 73-year-old male underwent endovascular repair of a 6-cm infrarenal aortic aneurysm using a Medtronic Endurant device (Medtronic, Wrexham, Wales, UK). Surveillance computed tomography scans demonstrated type 2 endoleak from a lumbar artery and the sac size increased by >1 cm. Subsequent attempts at lumbar artery branch embolisation were unsuccessful and percutaneous transabdominal glue injections were ineffective. The aneurysm enlarged to 7.7 cm at 27 months. The patient then developed episodes of abdominal pain; transperitoneal mini-laparotomy, sacotomy, and oversew of the lumbar artery were performed, leaving the stent graft undisturbed. After surgery, the sac diameter measured 5.3 cm. The patient remains well 5 years after the sacotomy procedure. * Corresponding author. E-mail address: Mohamed.elkawafi@hotmail.com (M. Elkawafi). 1078-5884/Ó 2016 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejvs.2016.10.027