Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair

Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair

Eur J Vasc Endovasc Surg (xxxx) xxx, xxx Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular A...

2MB Sizes 0 Downloads 3 Views

Eur J Vasc Endovasc Surg (xxxx) xxx, xxx

Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair Rosamaria Tricarico a,y, Roger Tran-Son-Tay Scott A. Berceli a,c,d, Yong He c,*

a,b

, Liza Laquian c, Salvatore T. Scali

c,d

, Teng-Chun Lee

e,z

, Adam W. Beck

c,x

,

a

Department of Biomedical Engineering, University of Florida, Gainesville, FL, USA Department of Mechanical and Aerospace Engineering, University of Florida, Gainesville, FL, USA c Division of Vascular Surgery and Endovascular Therapy, University of Florida, Gainesville, FL, USA d North Florida/South Georgia Veterans Health System, Gainesville, FL, USA e Division of Thoracic and Cardiovascular Surgery, University of Florida, Gainesville, FL, USA b

WHAT THIS PAPER ADDS Technologies allowing endografting in the aortic arch are in active clinical trials with various configurations of branch stent graft design. A systematic computational fluid dynamic analysis was carried out to quantitatively evaluate the haemodynamic characteristics of stenting within the left subclavian artery (LSA). The results showed that a shorter extension of the LSA branch stent graft into the aortic lumen up to 5 mm has a smaller risk of potential thrombus generation; a longer extension could lead to thrombus formation impacting outcomes. These findings may help optimise the deployment and design of the stent graft system for treating aortic pathologies involving arch branches.

Objective: Branched stent grafts represent a viable option for left subclavian artery (LSA) revascularisation in patients treated by thoracic endovascular aortic repair (TEVAR) for Zone 2 lesions. This study investigated the haemodynamic performance of different LSA branched stent graft configurations as potential determinants of thrombotic and stroke risks. Methods: A three dimensional aortic arch geometry extracted from post-operative computed tomography images of a TEVAR patient using a single LSA branched aortic endograft was modified in silico to obtain ten potential LSA branched stent graft configurations: five down facing (0e5 e 10 mm aortic protrusion with 10e12 mm internal diameter), four curved (30e60 with antegrade/retrograde orientation), and one LSA orifice misalignment. The 0 mm down facing stent graft was considered base configuration. Computational fluid dynamic analyses were performed to identify differences in pressure, energy, and wall shear stress (WSS) based parameters. Results: Total pressure drop and energy loss variations among configurations were not greater than 5 mmHg (6% of mean arterial pressure) and 5.7 mW (0.7% of cardiac power), respectively. Protrusions up to 5 mm created clinically insignificant flow disturbances. However, stent graft protrusions further into the aortic lumen created more complex haemodynamics, characterised by larger energy loss and more prominent flow recirculation. Protrusion greater than 5 mm into the lumen was associated with larger areas of elevated maximum WSS (>20 Pa) along the outer surface of the branched stent graft. Conclusion: Arterial haemodynamic characteristics are affected by LSA branched stent graft configurations, with pressure drops and energy losses likely to be clinically insignificant. The length of the stent graft protrusion into the aortic lumen generated the largest haemodynamic variations in the aortic system. Protrusions up to 5 mm have smaller risk of potential thrombus generation. Conversely, larger protrusions into the aortic lumen showed more disturbed haemodynamics, suggesting a greater risk of potential thrombus formation, which may be clinically important over time. Keywords: Aortic arch aneurysm, Branched endograft, Computational fluid dynamics, Haemodynamics, Left subclavian artery stent graft, Thoracic endovascular aortic repair Article history: Received 1 October 2018, Accepted 9 June 2019, Available online XXX Ó 2019 European Society for Vascular Surgery. Published by Elsevier B.V. All rights reserved.

y Centre for Regenerative Medicine of the Research Institute at Nationwide Children’s Hospital, Columbus, OH, USA.z Division of Cardiothoracic Surgery, University of California, San Francisco, CA, USA.x Division of Vascular and Endovascular Therapy, University of Alabama at Birmingham, Birmingham, AL, USA. * Corresponding author. University of Florida and Malcom Randall VAMC, PO Box 100128, Gainesville, FL, USA. E-mail address: [email protected]fl.edu (Yong He). 1078-5884/Ó 2019 European Society for Vascular Surgery. Published by Elsevier B.V. All rights reserved. https://doi.org/10.1016/j.ejvs.2019.06.028

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

2

Rosamaria Tricarico et al.

INTRODUCTION Minimally invasive thoracic endovascular aortic repairs (TEVARs) have been used increasingly to treat aortic diseases involving the aortic arch. TEVAR landing in Zone 2 of the Ishimaru aortic map1 is necessary for achieving an adequate proximal seal in up to 40% of TEVAR procedures.2 Left subclavian artery (LSA) coverage has been associated with stroke, arm, and spinal cord ischaemia; therefore, LSA revascularisation in selective TEVAR cases is recommended.3e5 Fenestrated/branched and parallel stent grafts are strategies used increasingly for LSA revascularisation.4 Currently fenestrated/branched stent grafts are available as custom made devices only, while off the shelf fenestrated/branched stent graft solutions are being developed. Parallel grafts are available off the shelf and can be used for emergency or urgent repair. Stroke remains one of the major concerns of endovascular aortic arch repair.6 Most reported strokes occur intraoperatively or during the early post-operative period and are due to emboli released during delivery and manipulation of the stent grafts, air emboli released from the delivery system, and coverage of the branch vessels. Stent graft deployment is a trade off between achieving device stability and reducing stroke risk. Given the distorted three dimensional geometry that is inherent in a severely diseased aorta requiring arch TEVAR, optimal placement of branched stent grafts using guidewires and catheters in a dynamic constrained environment is very challenging. Although optimal LSA branched stent graft positioning as determined in pre-procedure planning would include only

slight protrusion into the aortic lumen, suboptimal placement, such as longer protrusion of the LSA branch stent graft into the aortic lumen, is common in clinical practice. During LSA revascularisation, the branch orifice on the aortic stent graft may also misalign. In addition to different placements, the branch stent graft can be designed to have different levels of antegrade or retrograde curvature. The impact of these geometric variations on the aortic and LSA haemodynamics is unknown. Although strokes after the peri-operative period are less frequent in the reports concerning mainly the early results of arch TEVAR, thrombus formation on the thrombogenic stent graft material is common.7 Severe haemodynamic conditions induced by certain stent graft configurations may promote postimplantation thrombus formation on the branch or aortic stent graft and/or reduce distal blood flow.8 Based on a realistic aortic arch geometry and measured flow and pressure boundary conditions, this report describes the haemodynamics of the LSA branched stent graft and the adjacent aorta, through systematic computational fluid dynamics (CFD) exploration of a variety of stent graft configurations including potential geometries that might be engineered in future designs of the LSA branch stent graft. Total pressure drop (clinically referred to “pressure gradient”) and energy loss were calculated,9,10 respectively, as local and global measures of haemodynamic impact of the different stent graft configurations on LSA perfusion, together with wall shear stress (WSS) based parameters,11e 16 to assess potential determinants of thrombotic/stroke risk after arch TEVAR.

down-facing

A

B

C

Base configuration

Straight configurations

Oversized straight configurations

.

. .

. .

0 mm: S-0

5 mm: S-5 10 mm: S-10

misalignment

curved

5 mm: O-5 10 mm: O-10

D

E

Anterograde configuration

Retrograde configuration

.

.

.

30 degrees: C-30°-A 60 degrees: C-60°-A

.

30 degrees: C-30°-R 60 degrees: C-60°-R

F

LSA orifice misalignment

.

0 mm: M-0

Figure 1. Schematics of the 10 in silico modified configurations. Five straight (down facing) stent graft configurations, four curved configurations, and one configuration with misalignment of the LSA orifice were created from the patient specific anatomy. (A) Branch stent graft with a base diameter (10 mm) and 0 mm protrusion into the aortic lumen. (B) Branch stent grafts with base diameter and two different intra-aortic protrusion lengths (5 and 10 mm). (C) Branch stent grafts with an oversized diameter (12 mm) and two different intra-aortic protrusion lengths (5 and 10 mm). (D) Branch stent grafts with base diameter, two different antegrade curvatures (30 and 60 ) corresponding to 8 and 15 mm intra-aortic protrusions. (E) Branch stent graft with base diameter, two different retrograde curvatures (30 and 60 ), corresponding to 8 and 15 mm intra-aortic protrusions. (F) Branch stent graft with base diameter, misalignment between the aortic fenestration and LSA orifice (30 clockwise rotation), and 0 mm protrusion into the aortic lumen. LSA ¼ left subclavian artery.

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

Haemodynamics of LSA Stent Graft Configurations in Context of TEVAR

MATERIALS AND METHODS Generation of realistic anatomy The reference anatomy used for subsequent in silico modelling was obtained from a 71 year old female who underwent repair of a thoracic aortic aneurysm with an LSA branched stent graft. The aneurysm was repaired using an aortic stent graft (34  150 mm, diameter and length respectively) with proximal attachment in Zone 2 of the aortic arch, distal to the orifice of the left common carotid artery. Perfusion of the LSA was maintained via a self expanding stent graft (12  40 mm). Computed tomography (CT) images used for creation of the base geometry were obtained one week following repair (CT scanner;  Aquilion PRIME, Toshiba, Otawara, Japan); with in-plane imaging resolution 0.78  0.78 mm2; slice thickness 2 mm). Approval for this study was obtained from the University of Florida Institutional Review Board (Gainesville, FL), and informed patient consent was obtained. Computational modelling framework The patient’s post-operative thoracic aortic geometry was reconstructed from cross sectional CT images using Mimics 18.0 (Materialize, Leuven, Belgium). The aortic model was smoothed in Geomagic Studio (3D Systems, Morrisville, NC, USA) to reduce noise and surface irregularities and the arterial outlets were extended by 10 times their diameters in Vascular Modeling Toolkit (VMTK) to minimise the influence of the outlet boundaries. In silico modification. To encompass potential variation of device placements and designs that might be of interest, three categories of LSA branched stent graft were created (down facing, curved, and misaligned), for a total of 10 configurations obtained by varying length, diameter, orientation, and curvature of the intra-aortic extension of the LSA branched stent graft (Fig. 1). (i) Five straight placements (downward facing orifice) with variations of intra-aortic length (denoted by S followed by a number indicating the length of the stent graft extending into the aortic lumen in millimetres, Fig. 1A and B) or a larger internal diameter (denoted by O followed by the length of the stent graft extension, Fig. 1C); (ii) four curved device configurations with antegrade (upstream facing orifice) or retrograde orientation (downstream facing orifice) (denoted by C, followed by the amount of curvature in degree, and A for anterograde or R for retrograde orientation, e.g. C-30 A, Fig. 1D, and C-30 -R, Fig. 1E); and (iii) one aortic stent graft opening LSA orifice misalignment, leading to an S shaped branch stent graft geometry (M0, Fig. 1F). The shortest down facing configuration was assumed as the base configuration (S-0, Fig. 1A). The stent graft surface was assumed to be smooth, with a thickness of 0.7 mm. Its length varied from 30 to 45 mm, depending on the extension of the LSA device inside the aortic lumen; additional details regarding the geometric characteristics of the stent graft configurations are presented in Fig. S1. All configurations were generated by modifying the stereolithography

3

(STL) files in SpaceClaim Direct Modeler (ANSYS Inc., Canonsburg, PA, USA). Mesh generation. Tetrahedral meshes were created using ANSYS ICEM CFD R17.1. Four boundary layers and a total of 2.2 million mesh elements circa (minimum and maximum element size of 0.4 mm and 2 mm, respectively) were generated for all 10 models. The mesh size was chosen to be sufficient for resolving WSS based parameters within 5% accuracy within the LSA. CFD simulations. Using a previously published modelling algorithms,17,18 haemodynamics were simulated by numerically solving the continuity and NaviereStokes momentum equations for a pulsatile blood flow using ANSYS Fluent. Patient specific velocity profiles, derived from Doppler ultrasound imaging were adopted as outlet boundary conditions of the bilateral common carotid, subclavian, and vertebral arteries. The pressure waveform measured intra-operatively in the descending aorta at diaphragm level was used as the aortic outlet boundary condition. Details of the boundary conditions are presented in Fig. S2. Blood was treated as an incompressible, homogeneous, and Newtonian fluid, under a laminar regimen, with values of density and dynamic viscosity of 1050 kg/m3 and 0.0035 Pa s, respectively. The same boundary conditions were imposed on all simulations for comparison purposes. Additional CFD simulation details are included in the Supplementary material. Data analysis. CFD results were analysed using Tecplot 360 (Tecplot Inc., Bellevue, WA, USA). In order to highlight the major effects of the different stent graft placements, data were presented on the stented portion of proximal LSA (30 mm from the native orifice) and on the centreline along the aortic arch (200 mm from the aortic inlet, Fig. S2). Centrelines along the arterial tree were computed in VMTK. The total pressure, defined as the sum of dynamic and static pressures, was extracted from the LSA and aortic arch at peak systole; the pressure drop was quantified as the difference in total pressure between the aortic inlet and arterial portions distal to areas of interest or to the location of the descending thoracic aorta 200 mm from the aortic inlet to evaluate the differences in flow resistance caused by the branch stent graft protrusions into the aortic lumen. The total energy loss9 was calculated as the difference in energy between the aortic inlet and the seven outlets, in one cardiac cycle (T ¼ 0.88 s). Additional details are reported in the energy loss calculation within the Supplementary material. Both high and low shear flow can be thrombogenic. A selection of WSS based parameters was analysed to identify potentially thrombogenic areas. Specifically, maximum WSS (maxWSS),11 time averaged WSS (TAWSS), oscillatory shear index (OSI),12 and relative resident time (RRT)13 were extracted. A maxWSS greater than 18 Pa has been identified as an indicator of an area at risk of platelet activation.11,14 Conversely, RRT is useful for locating areas of high OSI (>0.25)15 and low TAWSS (<0.4 Pa),16 such as areas of low

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

4

Rosamaria Tricarico et al.

Table 1. Total pressure drops at peak systolic pressure across the stented LSA and the aortic arch Configuration

S-0 (base) S-5 S-10 O-5 O-10 M-0 C-30 -A C-60 -A C-30 -R C-60 -R

Systolic pressure drop e mmHg LSA stent graft

Aortic arch

3.9 4.7 5.6 4.2 5.4 4.1 2.5 1.0 5.3 3.9

2.0 2.3 2.6 2.3 2.9 2.0 2.5 3.4 2.5 2.6

Refer to Fig. 1 for the definitions of different stent graft configurations. LSA ¼ left subclavian artery.

velocity flow recirculation, resulting in high RRT (>5 1/Pa). To capture different aspects of the local haemodynamics and identify major areas of disturbed flow, each parameter was averaged across the circumferences orthogonal to the centreline of the LSA stent graft (every 0.5 mm) and distribution maps of the external surface of the stent graft were investigated. Results are presented as mean  standard deviation.

on these results, no analysed configurations were expected to undermine blood flow distal to the LSA. The peak systolic pressure along the aortic centreline from the aortic inlet to descending thoracic aorta is shown in Fig. S3. The results demonstrate that longer branch stent grafts caused a higher proximal pressure load, due to the increased intra-aortic resistance caused by the pronounced protrusions into the lumen. The aortic flow recirculation distal to the LSA orifice (Fig. 2D and E) might have contributed to the large drop in pressure just distal to this region (about 90 mm from the aortic inlet). Nonetheless, the largest and smallest aortic inlet pressure, observed in C60 -A and S-0 base configurations, respectively, differed only by 1.4 mm Hg (Table 1). The energy loss was compared among the different configurations (Fig. 3). The only stent graft configuration that presented a total energy loss comparable to the S-0 base configuration (7.8 mJ) was the M0 configuration (8.0 mJ). The C-60 -A configuration showed the highest value of energy loss, 12.8 mJ (þ64% of base configuration). These findings demonstrate that longer intra-aortic stent graft protrusions generate greater energy losses because of the larger aortic flow obstruction and resultant flow recirculation. Despite the differences from the base configuration, variations in energy losses among the various configurations were less than 5 mJ (5.7 mW), which represented 0.7% of the estimated cardiac power at rest (calculated as cardiac output multiplied by the mean arterial pressure).

RESULTS Pressure and energy analyses

WSS based parameter analysis

Table 1 summarises the pressure drops at peak systole from the aortic inlet to the distal end of the LSA stent graft. C60 -A had the lowest pressure drop along the LSA (74% lower than the base configuration), proving that anterograde stent graft orientation reduces the resistance across the LSA orifice. Conversely, S-10, O-10 and C-30 -R had the highest pressure drops (36%e44% higher than the base configuration). For both antegrade and retrograde configurations, the C-60 case had a smaller pressure drop than the corresponding C-30 case. Despite the differences from the base configuration, variations of pressure drops among the analysed configurations were <5 mmHg, which represented 6% of the measured mean arterial pressure. Examination of the spatial differences in peak systolic pressure demonstrated unique patterns along the length of the LSA stent graft (Fig. 2). All straight configurations showed a pressure peak in the area with the largest flow recirculation (Fig. 2A, C), which was evident in an examination of the dynamic velocity streamlines (Supplementary material). The 60 configurations presented fairly constant pressures along the LSA (Fig. 2B), signifying a reduced resistance to flow as demonstrated by a more organised flow pattern inside the stent graft. Low velocity blood recirculation was observed inside these configurations (Fig. 2D and E). In contrast, the M0 and 30 curved configurations, both antegrade and retrograde, showed pressure patterns similar to the straight configurations. Based

Differences in maximum WSS within the stented portion of the LSA were relatively large between some of the configurations examined (Fig. 4A). The proximal portion of M0 demonstrated the highest maxWSS (16.3 Pa) and the highest average across the stent graft (6.5  3.6 Pa), while the lowest values were found within the 60 curved configurations (2.3  0.6 Pa on average). A similar pattern was observed for TAWSS (Fig. 4B). The OSI was relatively constant across all stent graft configurations (0.14  0.05 on average, Fig. 4C). In terms of RRT (Fig. 4D), the C-60 -A and C-60 -R stent graft configurations demonstrated the highest averages (3.7  1.4 and 3.7  1.0 1/Pa, respectively), with these geometries exhibiting larger local regions of RRT > 5 1/Pa. The analysis of velocity magnitude coloured streamlines (Supplementary material) demonstrated a spiral high velocity flow during systole inside the first 20 mm of all LSA stent graft configurations, except for both 60 curved geometries where a less disturbed systolic flow was observed. Conversely, these latter configurations exhibited the most prominent regions of low velocity recirculation inside the LSA stent graft. Furthermore, greater high velocity flow recirculation was observed in the aortic arch of the longer branch stent graft protrusion into the lumen, due to the larger obstruction to aortic flow. The spatial distributions of maxWSS, RRT, TAWSS, and OSI were evaluated on the external surface of the LSA stent graft, as it protrudes into the aortic lumen and may be a

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

Haemodynamics of LSA Stent Graft Configurations in Context of TEVAR

S-0 S-5 S-10 O-5 O-10

122 120 118 116 114 -15

-5

5 15 LSA Centerline – mm

C

25

B

M-0 C-30°-A

Peak systolic pressure – mmHg

Peak systolic pressure – mmHg

A

5

D

C-60°-A C-30°-R C-60°-R

122 120 118 116 114 -15

-5

5 15 LSA Centerline – mm

25

Velocity [m/s] 0.30

E

0.26 0.22 0.18 0.14 0.10 0.06 S-0

C-60°-A

C-60°-R

0.02

Figure 2. Peak systolic pressures and velocity streamlines inside the proximal LSA. (A) Straight (down facing) configurations. (B) Curved and orifice misalignment configurations. Negative abscissae represent the portion of the LSA stent graft extended into the aortic lumen. (CeE) Streamlines showing flow patterns at peak systole for the S-0 base (C), C-60 -A (D), C-60 -R (E) configurations. The streamlines are colour coded by the velocity magnitude. Refer to Fig. 1 for the definitions of different stent graft configurations. LSA ¼ left subclavian artery.

region with a high risk in thrombus formation (Fig. 5 and Fig. S4 e S6). Larger areas of elevated maxWSS (>20 Pa) along the outer surface of the stent graft were associated with longer stent graft configurations (S-10, O-10, C-60-A, and C-60-R); smaller areas of elevated maxWSS were observed on the configurations with intermediate extension (S-5, O-5, C-30-A, and C-30-R).

S-0 S-5 S-10 O-5 O-10

DISCUSSION

M-0

Clinical implications of findings

C-30°-A C-60°-A C-30°-R C-60°-R 5

7

9 11 Total energy loss – mJ

13

15

Figure 3. Total energy loss in one cardiac cycle across the aortic arch and proximal branch arteries for the ten aortic configurations. Refer to Fig. 1 for the definitions of different stent graft configurations.

Motivated by ongoing advances and expanded clinical use of arch TEVAR, this analysis sought to examine the haemodynamic performance of different LSA branched stent graft configurations with TEVAR landing in aortic arch Zone 2. In summary, the investigated configurations did not show critically large changes in maximum pressure drop (across the LSA or aortic arch) and energy loss, suggesting limited impact on distal perfusion. Nonetheless, the results demonstrated that, despite antegrade oriented configurations reducing local LSA resistance, long protrusion into the aortic lumen creates flow disturbances that have a major haemodynamic impact on the aortic system. Associated with longer protrusions into the lumen are flow patterns

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

6

Rosamaria Tricarico et al.

S-0

5 0 -15

OSI

-5

5

15

25

4 3

O-10

2 1 5

15

5 0 -15

-5

5

15

25

-15

-5

5

15

25

5

15

25

4

25

3 2 1

0.4

0.4

0.3

0.3

0.2

M-0

-15

-5

5

15

25

C-30°-R

8

C-60°-R

0

0 -15

C-60°-A

12

4

0.2 0.1

C-30°-A

0

RRT – 1/Pa

10

0 -5

0.1

D

15

S-10 O-5

0 -15

C

S-5

TaWSS – Pa

10

max WSS – Pa

15

OSI

TaWSS – Pa

B

20

20

-5

12 RRT – 1/Pa

max WSS – Pa

A

8 4 0

-15

-5

5

15

25

Distance from LSA orifice towards LSA – mm

-15 -5 5 15 25 Distance from LSA orifice towards LSA – mm

Figure 4. Various wall shear stress (WSS) based parameters within the stented portion of LSA: maximum (max) WSS (A); time averaged (Ta) WSS (B); oscillatory shear index (OSI) (C); relative residence time (RRT) (D). Negative values on the x-axis represent the portion of the LSA stent graft extended inside the aortic lumen. LSA ¼ left subclavian artery. Refer to Fig. 1 for the definitions of different stent graft configurations.

that induce an elevation in the maximum WSS on the outer surface of the branch stent graft, which might induce platelet activation and thrombus formation. Furthermore, the branch stent graft with a longer protrusion is subjected to a larger impact by the pulsatile blood flow and may be associated with a higher risk of stent graft migration and/or endoleaks. To address this particular issue, a side branch stent graft anchored to the upper surface of the aortic stent graft (“inner branched device”) might reduce flow disturbance in the aortic arch and decrease repetitive branch stent graft motion that could lead to material fatigue and endoleak.19 The study demonstrated that a modest diameter oversize of the branch stent graft has minor haemodynamics effects. From a haemodynamic standpoint, a protrusion into the aortic lumen around 5 mm seems acceptable; however, protrusion greater than 5 mm is often used to minimise the risk of stent graft disconnection. The misaligned stent graft presented haemodynamic characteristics similar to the base configuration (S-0), but should be avoided due to the inherent risk of deformation/ compression of the branch stent graft by the main aortic stent graft and aortic wall.

Non-invasive analysis of the rate of energy (i.e. power) loss has been proposed to quantify the effect of nonphysiological aortic flow. Patients with aortic stenosis presented differences in viscous power loss (over 10 mW) compared with patients with dilated aortas.10 A recent study analysed platelets activation potential and changes of blood perfusion into the cerebral arteries to evaluate four endovascular configurations involving the ascending aorta.8 They demonstrated that a higher risk of platelets activation and decreased cervical blood flow occurs when a single stent graft and multiple bypasses are utilised for blood perfusion of the supra-aortic branches. In this study, WSS related parameters were used to evaluate risk of thrombosis. However, the implication of high values of RRT in the complex anatomy of the aortic arch and with thrombogenic foreign materials (stent and graft) remains to be fully resolved. Another recent study compared five stent graft configurations deployed in the three aortic arch branches.20 Results showed that an antegrade curved LSA stent graft generates less resistance to flow than a retrograde configuration, allowing a 79% LSA flow rate increase. These outcomes demonstrate that the orientation of a branched stent

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

Haemodynamics of LSA Stent Graft Configurations in Context of TEVAR

S-5

S-10

O-5

O-10

From DAo

From AAo

S-0

7

maxWSS – Pa

C-30-A

C-60-A

C-30-R

C-60-R

From DAo

From AAo

Misalignment

20 15 10 5 0

Figure 5. Spatial distribution of the maximum wall shear stress (maxWSS) viewed from inside the aortic stent graft lumen, upstream (AAo) and downstream (DAo) of the left subclavian artery stent graft. AAo ¼ ascending aorta; DAo ¼ descending aorta. Refer to Fig. 1 for the definitions of different stent graft configurations.

graft can affect the haemodynamics. In this study, the length of the curved LSA stent graft protrusion was limited to 15 mm, which might be short enough not to generate critical pressure differences (and therefore resistance to flow) between antegrade and retrograde stent graft orientations. Several branched stent grafts have been developed to treat aortic arch lesions and are in investigational use. The Valiant Mona LSA21 (Medtronic, Fridley, MN, USA) and TAG22 (W.L. Gore Newark, DE, USA) thoracic stent grafts have a single branch configuration, intended to maintain LSA perfusion while achieving a more proximal seal. The Mona LSA stent graft does not have an anchoring component for the branch stent graft protrusion inside the aortic stent graft and is similar to the down facing configuration presented in the current report. In contrast, the TAG stent graft has an integrated retrograde inner tunnel for the insertion of the branch stent graft. The next generation of devices, the Zenith19 (Cook, Bloomington IN, USA) and Aortic RelayBranch23 (Terumo, Tokyo, Japan) branched thoracic stent grafts, are designed to engage more complex arch pathology using two integrated inner antegrade tunnels inside the aortic lumen, to maintain perfusion to the innominate and left common carotid arteries. The internal branches are anchored to the upper surface of the aortic

stent graft without protruding into the aortic lumen. Although such complex configurations were not examined in the current analysis, a systemic evaluation of the haemodynamics that accompany variations in device deployment may reveal clinically important effects not identified in the current single branch analysis. With continued development, new devices designed specifically for TEVAR in the aortic arch will be facilitated by advanced computational analyses using more realistic boundary conditions and mechanical properties, allowing optimisation of the haemodynamic performance of these future devices. Although off the shelf device configurations are ideal, allowing for readily available devices in emergency procedures, anatomy specific devices may be best to avoid haemodynamic conditions that might predispose to poorer outcomes. From this perspective, future endeavour should be to develop the pipeline for quickly manufacturing and delivering a long lasting anatomy matching stent graft system at a low cost. Given the inherent differences in patient anatomy, variations in the deployment of branch endografts are inevitable. To date, a systematic evaluation of the haemodynamic effects of branch protrusion, regional curvature, or rotational misalignment has not been performed. Focusing on the most common procedure performed for

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

8

Rosamaria Tricarico et al.

the treatment of aortic arch aneurysms, the current manuscript used in silico modelling coupled with computational fluid dynamic analysis to examine changes in the flow fields induced by deviations in subclavian branch graft deployment. Despite pronounced modifications in stent graft configuration, the associated pressure drop and wall shear stress profiles are relatively unaffected. While suboptimal deployment of branch grafts is unavoidable at times, these devices are highly tolerant of these variations, which are unlikely to result in clinically significant changes in distal perfusion.

CONFLICT OF INTEREST

Limitation and future implementation

APPENDIX A. SUPPLEMENTARY DATA

In silico modelling allows a first order approximation in the investigation of haemodynamics, where the boundary conditions play a fundamental role in the CFD analysis. In the absence of blood flow distribution variation among the 10 analysed stent graft configurations, the same outlet velocity profiles were imposed on all 10 configurations, aiming to compare the haemodynamic consequences under the expected patient specific physiological conditions. Blood flow redistribution among the aortic branches might affect the results. Recent analyses have suggested that haemodynamics and stent graft displacement forces differed depending on aortic arch geometries;24 therefore, the results are limited to a single patient’s anatomy (bovine arch). In the current analysis, aortic wall compliance and motion have been ignored. While a more detailed analysis using an integrated model is possible,25 it is unlikely that it would have a major impact on the relative differences observed in the current analysis. Only 10 configurations were analysed. Other important configurations might have been missed. To investigate the potential clinical significance of the haemodynamic findings further, the risk of thrombosis might be assessed by particle tracking tools.26 Additionally, dynamic motion of the LSA stent graft and aortic stent graft migration and endoleaks, due to inadequate fixation at both ends of the stent grafts, can be investigated through fluide structure interaction methods, although this requires much greater computational resources, and accurate acquisition of the mechanical properties of the aorta and branch arteries remains challenging. CONCLUSIONS The results showed that geometric changes of the LSA stent graft generate mild to moderate arterial haemodynamic variations, in terms of pressure drop and energy loss, which are not expected to undermine the distal flow rate. Comparison among the investigated stent graft configurations showed that a shorter LSA branch stent graft protrusion into the aortic lumen up to 5 mm carries a smaller risk of potential thrombus generation, while a longer protrusion into the aortic lumen was associated with augmented flow disturbance in the aortic lumen and a potentially higher risk of thrombus formation, and efforts should be taken to avoid these long protrusion cases.

None. FUNDING None. ACKNOWLEDGEMENTS This study was supported by CyBHOR (Cyber-Physical System for the Hospital Operating Room), an Industry/University Cooperative Research Centre (1266260) funded by the National Science Foundation.

Supplementary data to this article can be found online at https://doi.org/10.1016/j.ejvs.2019.06.028. REFERENCES 1 Mitchell RS, Ishimaru S, Ehrlich MP, Iwase T, Lauterjung L, Shimono T, et al. First International Summit on Thoracic Aortic Endografting: roundtable on thoracic aortic dissection as an indication for endografting. J Endovasc Ther 2002;9. II-98-II-105. 2 Peterson BG, Eskandari MK, Gleason TG, Morasch MD. Utility of left subclavian artery revascularization in association with endoluminal repair of acute and chronic thoracic aortic pathology. J Vasc Surg 2006;43:433e9. 3 Riambau V, Bockler D, Brunkwall J, Cao P, Chiesa R, Coppi G, et al. Editor’s choice - management of descending thoracic aorta diseases: clinical practice guidelines of the European Society for Vascular Surgery (ESVS). Eur J Vasc Endovasc Surg 2017;53:4e52. 4 Czerny M, Schmidli J, Bertoglio L, Carrel T, Chiesa R, Clough RE, et al. Current options and recommendations for the treatment of thoracic aortic pathologies involving the aortic arch: an expert consensus document of the European Association for CardioThoracic Surgery (EACTS) and the European Society for Vascular Surgery (ESVS). Eur J Cardio-thorac Surg 2019;55:133e62. 5 Matsumura JS, Lee WA, Mitchell RS, Farber MA, Murad MH, Lumsden AB, et al. The Society for Vascular Surgery practice guidelines: management of the left subclavian artery with thoracic endovascular aortic repair. J Vasc Surg 2009;50:1155e8. 6 Spear R, Haulon S, Ohki T, Tsilimparis N, Kanaoka Y, Milne CP, et al. Editor’s Choice - subsequent results for arch aneurysm repair with inner branched endografts. Eur J Vasc Endovasc Surg 2016;51:380e5. 7 Maleux G, Koolen M, Heye S, Heremans B, Nevelsteen A. Mural thrombotic deposits in abdominal aortic endografts are common and do not require additional treatment at short-term and midterm follow-up. J Vasc Interv Radiol 2008;19:1558e62. 8 van Bakel TM, Arthurs CJ, van Herwaarden JA, Moll FL, Eagle KA, Patel HJ, et al. A computational analysis of different endograft designs for Zone 0 aortic arch repair. Eur J Cardio-thorac Surg 2018;54:389e96. 9 Sia SF, Zhao X, Li R, Zhang Y, Chong W, He L, et al. Evaluation of the carotid artery stenosis based on minimization of mechanical energy loss of the blood flow. Proc Inst Mech Eng H 2016;230: 1051e8. 10 Barker AJ, van Ooij P, Bandi K, Garcia J, Albaghdadi M, McCarthy P, et al. Viscous energy loss in the presence of abnormal aortic flow. Magn Reson Med 2014;72:620e8. 11 Casa LDC, Ku DN. thrombus formation at high shear rates. Annu Rev Biomed Eng 2017;19:415e33. 12 He X, Ku DN. Pulsatile flow in the human left coronary artery bifurcation: average conditions. J Biomech Eng 1996;118:74e82. 13 Himburg HA, Grzybowski DM, Hazel AL, LaMack JA, Li XM, Friedman MH. Spatial comparison between wall shear stress

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028

Haemodynamics of LSA Stent Graft Configurations in Context of TEVAR

14 15

16 17

18

19

20

measures and porcine arterial endothelial permeability. Am J Physiol Heart Circ Physiol 2004;286:H1916e22. Jackson SP. The growing complexity of platelet aggregation. Blood 2007;109:5087e95. Gallo D, Lefieux A, Morganti S, Veneziani A, Reali A, Auricchio F, et al. A patient-specific follow up study of the impact of thoracic endovascular repair (TEVAR) on aortic anatomy and on postoperative hemodynamics. Comput Fluids 2016;141:54e61. Malek AM, Alper SL, Izumo S. Hemodynamic shear stress and its role in atherosclerosis. JAMA 1999;282:2035e42. He Y, Terry CM, Nguyen C, Berceli SA, Shiu YT, Cheung AK. Serial analysis of lumen geometry and hemodynamics in human arteriovenous fistula for hemodialysis using magnetic resonance imaging and computational fluid dynamics. J Biomech 2013;46:165e9. Tricarico R, He Y, Laquian L, Scali ST, Tran-Son-Tay R, Beck AW, et al. Hemodynamic and anatomic predictors of renovisceral stent-graft occlusion following chimney endovascular repair of juxtarenal aortic aneurysms. J Endovasc Ther 2017;24:880e8. Haulon S, Greenberg RK, Spear R, Eagleton M, Abraham C, Lioupis C, et al. Global experience with an inner branched arch endograft. J Thorac Cardiovasc Surg 2014;148:1709e16. Chiu TL, Tang AYS, Cheng SWK, Chow KW. Analysis of flow patterns on branched endografts for aortic arch aneurysms. Inform Med Unlocked 2018;13:62e70.

9

21 Roselli EE, Arko 3rd FR, Thompson mm. Results of the Valiant Mona LSA early feasibility study for descending thoracic aneurysms. J Vasc Surg 2015;62:1465e1471.e3. 22 Patel HJ, Dake MD, Bavaria JE, Singh MJ, Filinger M, Fischbein MP, et al. Branched endovascular therapy of the distal aortic arch: preliminary results of the feasibility multicenter trial of the Gore thoracic branch endoprosthesis. Ann Thorac Surg 2016;102:1190e8. 23 Ferrer C, Cao P, Coscarella C, Ferri M, Lovato L, Camparini S, et al. Italian RegIstry of doUble inner branch stent graft for arch PatHology (the TRIUmPH Registry). J Vasc Surg 2019;70:672e682. e1 https://doi.org/10.1016/j.jvs.2018.11.046. 24 Marrocco-Trischitta MM, van Bakel TM, Romarowski RM, de Beaufort HW, Conti M, van Herwaarden JA, et al. The modified arch landing areas nomenclature (MALAN) improves prediction of stent graft displacement forces: proof of concept by computational fluid dynamics modelling. Eur J Vasc Endovasc Surg 2018;55:584e92. 25 Jin S, Oshinski J, Giddens DP. Effects of wall motion and compliance on flow patterns in the ascending aorta. J Biomech Eng 2003;125:347e54. 26 Di Achille P, Tellides G, Figueroa CA, Humphrey JD. A haemodynamic predictor of intraluminal thrombus formation in abdominal aortic aneurysms. Proc Math Phys Eng Sci 2014;470: 20140163.

Please cite this article as: Tricarico R et al., Haemodynamics of Different Configurations of a Left Subclavian Artery Stent Graft for Thoracic Endovascular Aortic Repair, European Journal of Vascular and Endovascular Surgery, https://doi.org/10.1016/j.ejvs.2019.06.028