Hemodynamic Comparison of Stent-Grafts for the Treatment of Aortoiliac Occlusive Disease

Abstract

Purpose:

To compare the flow patterns and hemodynamics of the AFX stent-graft and the covered endovascular reconstruction of aortic bifurcation (CERAB) configuration using laser particle image velocimetry (PIV) experiments.

Materials and Methods:

Two anatomically realistic aortoiliac phantoms were constructed using polydimethylsiloxane polymer. An AFX stent-graft with a transparent cover made with a new method was inserted into one phantom. A CERAB configuration using Atrium’s Avanta V12 with transparent covers made with a previously established method was inserted into the other phantom, both modified stent-grafts were suitable for laser PIV, enabling visualization of the flow fields and quantification of time average wall shear stress (TAWSS), oscillatory shear index (OSI), and relative residence time (RRT).

Results:

Disturbed flow was observed at the bifurcation region of the AFX, especially at the end systolic velocity (ESV) time-point where recirculation was noticeable due to vortical flow. In contrast, predominantly unidirectional flow was observed at the CERAB bifurcation. These observations were confirmed by the quantified hemodynamic results from PIV analysis where mean TAWSS of 0.078 Pa (range: 0.009–0.242 Pa) was significantly lower in AFX as compared with 0.229 Pa (range: 0.013–0.906 Pa) for CERAB (p<0.001). Mean OSI of 0.318 (range: 0.123–0.496) in AFX was significantly higher than 0.252 (range: 0.055–0.472) in CERAB (p<0.001). Likewise, mean RRT of 180 Pa⁻¹ (range: 9–3603 Pa⁻¹) in AFX was also significantly higher than 88 Pa⁻¹ (range: 2–840 Pa⁻¹) in CERAB (p=0.0086).

Conclusion:

In this in vitro study, the flow pattern of a modified AFX stent-graft was found to be more disturbed especially at the end systolic phase, its hemodynamic outcomes less desirable than CERAB configuration.

Clinical Relevance:

While the AFX stent-graft has an advantage over the CERAB configuration in eliminating radial mismatch, and maintaining the anatomical bifurcation for future endovascular intervention, this in vitro study revealed that the associated lower TAWSS, higher OSI and RRT may predispose to thrombosis and are, thus, less desirable as compared to a CERAB configuration. Further investigation is warranted to confirm whether these findings translate into the clinical setting.

Keywords

aortoiliac occlusive disease aortoiliac bifurcation covered endovascular reconstruction of the aortoiliac bifurcation AFX endograft endovascular treatment particle image velocimetry time average wall shear stress oscillatory shear index relative residence time

Introduction

The covered endovascular reconstruction of the aortic bifurcation (CERAB) technique is a relatively new approach in treating extensive aortoiliac occlusive disease.^1,2 One of the advantages of the CERAB configuration as compared to the traditional kissing stent procedure is that it reduces radial mismatch, which is defined as the discrepancy between the stented lumen and the vessel lumen after stent placement.³ This leads to more favorable flow conditions. Initial clinical data for CERAB are encouraging, with 87.3% and 82.3% primary patency rate at 1- and 2-year follow-up in a group of 103 patients with 88% TransAtlantic Inter-Society Consensus II D lesions³ as compared with 91.5% primary patency rate at 2-year follow-up for open surgery.⁴ Taeymans et al⁵ further updated the CERAB results with a total of 130 patients showing that a primary patency of 82.0% was achieved at 3-year follow-up as compared with 90.2% for open surgery.⁶

More recently, it was suggested that the use of a unibody stent-graft, designed for aneurysm treatment, could have several advantages for the treatment of aortoiliac occlusive disease over CERAB. It preserves the aortic bifurcation and thus allows for future cross-over endovascular interventions and avoids limb competition in the distal aorta.^7
–10 Like CERAB, it also protects against potentially fatal aortoiliac rupture during dilatation of heavily calcified lesions. The AFX unibody stent-graft (Endologix, Inc, Irvine, CA, USA) is a thin-walled expanded polytetrafluoroethylene (ePTFE) sheath with an endoskeleton made of Elgiloy (cobalt-chromium alloy). The benefit of the AFX stent-graft lies in the fact that radial mismatch is completely eliminated because of the one-piece bifurcation design.⁷ The safety, efficacy, and early patency rates of the AFX stent-graft for treatment of aortoiliac occlusive disease were examined by several groups, totaling 128 patients.^7
–10 Primary patency rate for these studies were reported in the range of 80% to 100% at 1-year follow-up. The primary and secondary patency for the largest study with 90 patients were 78.8% and 100% at 3-year follow-up.

Even though both the off-label use of Advanta V12 in CERAB and AFX endograft are effective in treating extensive aortoiliac occlusive disease, the hemodynamic differences between them are currently unknown. As such, an in vitro study would provide new and valuable information for vascular surgeons and existing stent manufacturers to make further improvement in the treatment of aortoiliac occlusive disease. Particle image velocimetry (PIV) is a powerful experimental method to analyze hemodynamic parameters such as flow pattern, wall shear stress (WSS), oscillatory flow index (OSI) and relative residence time (RRT) in an in vitro setting. Laser PIV is the current gold standard for flow imaging,^11
–13 its use will help to better understand the influence of hemodynamic factors on blood flow in general and their impact on atherosclerosis in particular. A new method in making transparent cover for AFX stent-graft was introduced for the laser PIV study. The aim of this study is to test the hypothesis that the AFX stent-graft is superior to CERAB configuration where undisturbed flow patterns, physiological WSS 0.5 to 1.2 Pa, lower OSI and RRT are considered as better hemodynamic outcomes.¹⁴ Figure 1 illustrates the 2 approaches with Figure 1A and D showing the anterior view of AFX and CERAB, respectively.

Materials and Methods

Model Designs and Stent Deployment

Two identical aortoiliac bifurcation anatomies in which physiologic flow was reproduced, were used in this experimental study. A CERAB configuration with modified covered stents was deployed in one flow model and a modified AFX unibody stent in the other. The detailed production process for making these flow models has been previously described.¹⁵ The CERAB model was configured using 3 modified covered stents comprising two 8-mm and one 12-mm balloon expandable stents. These stents were derived from Advanta V12 stents (Getinge Group, Wayne, MA, USA) where the ePTFE graft covers were manually removed and subsequently individually coated with a transparent polyurethane cover (Tecoflex TPU clear; The Lubrizol Corporation, Wickliffe, OH, USA) using a dip coating process, which facilitated the visualization of flow patterns inside the flow model.^3,16 The polyurethane dipping process was not suitable to replace the ePTFE cover of the AFX stent because of the large spaces between the struts in its expanded configuration. The AFX stent cage was supplied by Endologix with 22-mm main body and 13-mm limbs. The cover with 18-mm body and 10-mm limb was 1 mm oversized compared with the wall-less, rigid polydimethylsiloxane (PDMS) phantom with 16-mm main body and 8-mm limb. The AFX ePTFE cover was replaced by low-density polyethylene (LDPE) film in the shape of 2 cylindrical-shaped grafts, which were stitched together to form the unibody. The covering material was on the outer layer, as with the original stent. Details of the AFX transparent cover-making process are found in Appendix A. A transparent LDPE cover was stitched onto the proximal frame the same way as the original, opaque ePTFE cover was secured on the frame of the AFX endograft. By replicating this feature, it enables the cover to move independently of the frame and we expected it to have the same sealing effect as the original AFX and allows for expansion throughout the cardiac cycle.

Experimental Flow Circuit

An experimental flow circuit as shown in Figure 2 was used to study and compare the hemodynamics in the AFX and CERAB. The flow loop was driven by a programmable hydraulic piston-driven pump (SuperPump) and its accompanying software Vivitest (Vivitro Labs Inc, Victoria, CA). A blood-mimicking fluid comprising water (44.07%), glycerol (34.52%), and urea (21.41%) was used to obtain a fluid with a refractive index optically matched to that of the PDMS model and physiologically matched to that of human blood (dynamic viscosity 4.2 mPa·s, density 1114 kg/m³).¹⁷ The blood-mimicking fluid was circulated through the flow loop by the SuperPump and a suprarenal flow profile mimicking human resting conditions with a mean flow rate of 1.6 L/min (or a stroke volume of 26.7 mL) and 60 beats per minute heart rate was generated by the Vivitest. A CORRI_FLOW M55 Coriolis flow meter (Bronkhorst High-Tech BV, Ruurlo, the Netherlands) was used to provide a real time display of the targeted suprarenal flow profile. Three UF8B ultrasound flow sensors (Cynergy3 Components Ltd, Wimborne, Dorset, UK) and 3 needle valves article number 136230 (Riegler Medical, Nieder-Ramstadt, Germany) were connected to left renal outlet, right renal outlet and the combined iliac outlets, respectively. A display unit provided real-time display of the flow rates of blood-mimicking fluid flowing through the ultrasound flow sensors. The needle valves were manually adjusted to achieve a mean pressure of 100 mm Hg and approximately 25% outflow, or 0.4 L/min, at the left and right renal outlet and approximately 50%, or 0.8 L/min, outflow at the combined iliac outlet. In addition, the 2 iliac outlets were connected to a single compliance chamber. A DS44 handheld sphygmomanometer (Welch Allyn, Skaneateles Falls, NY, USA) was used to set the distal peripheral pressure between 80 and 120 mm Hg.

Flow Visualization

In order to facilitate the visualization of the flow pattern inside the phantom, a continuous laser sheet was projected horizontally into the center of the stented lumen region of interest (ROI). The position of the phantom was manually adjusted in such a way that the laser sheet was projected onto the center of inflow and bifurcation ROIs as shown in Figure 3. Even though the ROIs for the AFX and CERAB are not identical due to minor differences in their configurations however, the scientific method of determining the velocity vector and wall shear stress using PIV method is the same in both cases. Therefore, it is not expected to have any direct influence on comparing the results. The laser sheet was slightly obscured by the renal branch at the inflow proximal section of AFX therefore its inflow ROI was shifted distally slightly as compared with that of the CERAB inflow. In addition, the bifurcation ROI for the AFX stent-graft was close to matching the native bifurcation because the configuration of the unibody AFX stent-graft was designed to mimic just that. However, the bifurcation ROI for the CERAB was located about 20 mm upstream. Rhodamine-coated fluorescent polymethyl methacrylate particles (size, 1–20 μm; density 1190 kg/m³) were suspended in the blood mimicking fluid. Digitized frames were collected by illuminating with a continuous-wave laser (5 W DPSS laser, 532 nm; Cohlibri, Lightline, Germany) and by recording using a high-speed camera (FASTCAM SA-Z; Photron Inc, West Wycombe, Buckinghamshire, UK). The images with 1024×1024 pixels resolution and 8 bit/pixel grayscale intensity were captured at a frequency of 2000 Hz. The camera was mounted with its optical axis perpendicular to the laser sheet. Two sets of images were captured for the CERAB-stented phantom at the proximal inflow and aortoiliac bifurcation and then repeated for the AFX-stented phantom.

Data Processing

Data processing was described previously.¹⁸ Since there are 2 regions of interest per stent configuration, data processing was performed on a total of 4 data sets. For each data set, 10 cardiac cycles were obtained and averaged to provide one averaged cycle per data set. Using an in-house developed MATLAB (version 2016b; The MathWorks Inc, Natick, MA, USA) script, the following hemodynamic parameters were calculated along the vessel wall: inflow, bifurcation right edge and bifurcation left edge were as shown in Figure 3. TAWSS, OSI, and RRT were computed according to the equations commonly used in hemodynamic studies.¹⁹

TAWSS is computed as the average WSS over a cardiac cycle where only the magnitude of the shear stress is taken into consideration. It is the average shear stress axially exerted on the vessel wall due to the blood flow motion. A value of 0.5 to 1.2 Pa is considered as physiological WSS in human arterial blood flow and values below and above this range are considered low and high WSS, respectively.^20,21 On the other hand, OSI takes into account the WSS direction with respect to the mean blood flow direction during the cardiac cycle. The OSI can vary from 0 to 0.5, 0 meaning no oscillation and 0.5 implying maximum oscillation. OSI has been studied as a relevant biomarker, for the assessment of vessel areas that are predisposed to develop atherosclerotic plaque.^22,23 RRT indicates the residence time of fluid particles spent at the vessel wall. Wall region with low WSS and elevated RRT may be prone to thrombus deposition.^24,25

Figure 4 shows the difference between flow profiles measured by flow sensor as compared with that determined with PIV method at the inflow over twenty cardiac cycles. The variation of flow rate is one magnitude larger using flow sensor measurement as compared with PIV method based on the statistical analysis of standard deviations.

Results

Inflow Pattern

Figure 5 shows the inflow patterns for AFX and CERAB at PSV, ESV and PDV time-points. At the PSV time-point, the maximum velocity was achieved as the blood-mimicking fluid accelerated into the phantom. A disturbed flow pattern was observed at the AFX inflow especially at the ESV time-point as shown in Figure 5B. Flow reversal was observed during the end-systolic phase and recirculation was noticeable in the AFX phantom (Video 1, online only; B). In contrast, unidirectional flow was observed in the CERAB inflow without noticeable disturbances (Video 2, online only; E).

Bifurcation Flow Pattern

Bifurcation flow pattern for AFX and CERAB at PSV, ESV and PDV time-points are shown in Figure 6. Disturbed flow was observed in the AFX stent in all 3 phases of the triphasic flow especially more pronounced during the end-systolic phase (Video 3, online only; A-6C). In contrast, only slightly disturbed flow was observed in the CERAB configuration during the end-systolic phase (Video 4, online only; E).

Hemodynamic Comparison

Table 1 summarizes the hemodynamic comparison (TAWSS, OSI, and RRT) of AFX and CERAB at the inflow section and the bifurcation. The mean and range for each set of TAWSS, OSI and RRT data were determined and presented in the table. In addition, the p-values for each pairwise Student’s t test between AFX stent-graft and CERAB configuration are also included in the table. The quantified hemodynamic parameters are in agreement with the observation of more disturbed flow in the bifurcation of the AFX stent-graft as compared with CERAB configuration. More details of hemodynamic plots of TAWSS, OSI, and RRT along the vessel wall can be found in Appendix B.

Table 1.

Time-Averaged Wall Shear Stress (TAWSS), Oscillatory Shear Index (OSI), and Relative Residence Time (RRT) Results at Inflow and Bifurcation for AFX Stent-Graft and CERAB Configuration^a.

Region of Interest		Stent	TAWSS (Pa)		p	OSI		p	(RRT (Pa⁻¹)		p
Region of Interest		Stent	Mean	Range	p	Mean	Range	p	Mean	Range	p
Inflow		AFX (n = 20)	0.185	0.036–0.296	0.0048	0.316	0.207–0.415	0.283	55	8–342	0.646
Inflow		CERAB (n = 22)	0.116	0.064–0.175	0.0048	0.340	0.304–0.428	0.283	44	15–158	0.646
Bifurcation	Right edge	AFX (n = 114)	0.074	0.013–0.186	0.0019	0.307	0.123–0.496	4.6×10⁻⁵	134	13–1744	0.35
	Right edge	CERAB (n = 61)	0.145	0.013–0.650	0.0019	0.237	0.055–0.472	4.6×10⁻⁵	103	2–840	0.35
	Left edge	AFX (n = 57)	0.088	0.009–0.242	2.59×10⁻⁸	0.339	0.217–0.482	0.93	273	9–3,603	0.00044
	Left edge	CERAB (n = 12)	0.695	0.448–0.906	2.59×10⁻⁸	0.337	0.250–0.387	5	273	3–7	0.00044
	Total	AFX (n = 171)	0.078	0.009–0.242	5.53×10⁻⁶	0.318	0.123–0.496	2.16×10⁻⁵	180	9–3603	0.0086
	Total	CERAB (n = 73)	0.229	0.013–0.906	5.53×10⁻⁶	0.252	0.055–0.472	2.16×10⁻⁵	88	2–840	0.0086

Abbreviation: CERAB, covered endovascular reconstruction of aortic bifurcation.

“n” is the number of spatial points available at the regions of interest.

Discussion

In the current study, a method in making transparent cover for AFX stent-graft was introduced. Laser PIV experiments were performed in AFX and CERAB flow phantoms. The images obtained from the experiments were processed to analyze flow patterns and compare TAWSS, OSI, and RRT. A more disturbed and oscillatory flow was observed in the flow pattern of the bifurcation region in the AFX stent-graft as compared with the CERAB covered stent configuration that showed an undisturbed flow pattern. This result was supported by the quantified hemodynamic parameters where TAWSS was lower in AFX and RRT was higher in AFX stent-graft as compared with the CERAB configuration. This was contrary to the hypothesis of our study. The AFX stent-graft designed for treatment of aneurysmal disease, but used by some surgeons for treatment of occlusive disease, has the advantage over a CERAB configuration because radial mismatch between the stented lumen and vessel lumen is eliminated due to its unibody design. Moreover, the aortic bifurcation is preserved permitting repeated long-term endovascular access. However, from flow pattern and hemodynamic perspective, this study showed that CERAB configuration potentially has the advantage of preventing thrombosis due to less disturbed flow, higher TAWSS, lower OSI and RRT as compared with AFX. WSS below physiological level has been shown to be predictive of plaque development based on the hypothesis that low shear stress promote modification of mass transport of atherogenic substances between the lumen and the vessel wall to cause atherosclerosis. Low and oscillating WSS is associated with the development of thrombosis.¹⁴

TAWSS and OSI in this study ranged from 0.009 to 0.906 Pa and 0.1 to 0.496, respectively as compared with mean inflow WSS at PSV, ESV, and PDV time-points of 1 Pa, −0.3 Pa, and 0.4 Pa; and OSI ranged from 01 to 0.2 at the bifurcation in the previous similar hemodynamic study,³ which is in line with this current study. In addition, WSS values ranging from 0 to 0.80 Pa and OSI values ranging from 0.01 to 0.49 were obtained in the CERAB results involving WSS study,²⁶ also in agreement with our current study. There was a slightly more disturbed flow at ESV at the CERAB bifurcation in the current study compared with what was observed in the control model of the previous hemodynamic comparison study.³ The higher WSS values in the CERAB configuration in the bifurcation could also be attributed to the fact that the flow transition from the bifurcation in the distal aorta to the iliac branches is more orderly with higher velocity near the boundary layer along the vessel wall, which might provide protection from thrombosis. It is important to note that the TAWSS observed in both the AFX stent-graft and the CERAB configuration at the inflow and at the bifurcation in this in vitro study were lower than the physiological WSS value of 0.5 to 1.2 Pa in human arteries.^20,22,23 In addition, the maximum value of TAWSS for CERAB at inflow was 0.906 Pa at the bifurcation which was significantly higher than 0.242 Pa for AFX therefore more protective of thrombosis for CERAB in the bifurcation region (p=5.53×10⁻⁶). Since blood is not in direct contact with endothelial cells, due to the ePTFE graft material used in the covers of both AFX stent-graft and Atrium V12 stents, the graft material provides some protection even though TAWSS might be below physiological values of 0.5 Pa.²⁰ However, platelet deposition and neointimal hyperplasia is a risk factor for synthetic graft. Furthermore, graft thrombosis is possible even with use of cover.²²

There are several limitations to our study: First, this in vitro study used polyurethane and LDPE for the modified transparent covers with coefficient of friction (0.12–0.70) and (0.20–2.5), respectively, which are slightly higher than that of the original opaque ePTFE covers (0.05–0.10) therefore might influence the WSS. Second, the PIV method is a two-dimensional analysis of a 3-dimensional flow phenomenon therefore only one slice of the 2-dimensional flow pattern centered on the lumen could be recorded for each PIV analysis. Multiple slices would be required to get a more complete picture of 3-dimensional flow across the lumen as a whole, which could reveal areas of stagnant flow with low TAWSS in the helical CERAB configuration. Third, the laser sheet was obscured by the renal branch of the AFX phantom therefore unable to capture PIV images upstream of the AFX stent-graft. In the existing in vitro setup, a suprarenal flow profile is prescribed at the inlet and a triphasic flow profile at the infrarenal aorta is achieved by adjusting resistances of renal and iliac arteries in addition to the adjustment of the compliance of the iliac arteries. The renal arteries can be eliminated if the pump can provide an infrarenal flow profile. An alternative approach would be to apply an infrarenal flow profile at the inlet directly using a novel roller pump,²⁷ or a similar pump, in order to eliminate the renal branches in the flow model as to allow an unobstructed laser sheet entry into the model to produce complete AFX inflow. Finally, the proximity of the AFX to the renal arteries could explain part of the recirculations in the inflow,²⁶ which could generate strong flow disturbances that are less normalized at the AFX relative to the CERAB. However, clinically AFX is likely to be deployed nearer to the renal arteries than CERAB, which remains low.

The aortoiliac model used is considered rigid, which in many cases can be considered appropriate given the often highly calcified lesions in patients with aortoiliac occlusive disease. The AFX graft and CERAB configuration have different properties in terms of flexibility, which may also influence patency results. However, this was outside the scope of the current study. The disturbed and oscillatory flow might have been caused by the inner skeleton of the AFX stent-graft, protruding into the vessel lumen. In addition, the infolding of the graft material could also be a factor as the stent-graft was slightly oversized. We followed the manufacturer recommendation of 10% to 20% oversize where the stent graft was 1 mm oversized as compared with the lumen diameter. In the current study a transparent AFX device was designed, in order to enable laser PIV, considered the gold standard of flow imaging. Future studies using the original AFX, in combination with ultrasound-PIV might confirm whether infolding of LDPE graft material contributed to the disturbed flow. Another future study using computational fluid dynamics (CFD) methods to simulate the experimental setup can be performed to predict the hemodynamics of the 3-dimensional flow model. When the results of this study are combined with previous literature,³ the hemodynamic consequences of bare metal stent, covered kissing stents, CERAB configuration and the AFX stent-graft for treating aortoiliac occlusive disease can be compared. From a hemodynamic perspective, CERAB seems to have a slightly better performance compared with the other 3 types of stents. The AFX stent-graft has the geometric advantage of eliminating radial mismatch but this is counteracted by disturbed flow in the bifurcation region. Therefore, it might be beneficial to investigate design improvement such as an inner ePTFE graft to smoothen the protruding strut links. This approach potentially reduces disturbed flow if it is to be considered a definitive treatment for occlusive disease rather than aneurysm for which it was designed. In the present study, we were only interested in comparing the hemodynamics of the 2 stent configurations. As such a negative control does not add value and we did not include a negative control without a stent graft.

Conclusion

In this in vitro study, it was found that the AFX may be less desirable as compared to the CERAB configuration with regards to flow changes, and when used for treatment of aortoiliac occlusive disease. Further investigations are warranted to confirm whether the in vitro hemodynamic results identified in this study are translated into the clinical setting and correlate with an increased incidence of thrombosis.

Footnotes

Appendix A

Fashioning Process for AFX Transparent Cover.

Step	Description	Process
1	Remove ePTFE cover from the original AFX stent-graft.
2	Cut LDPE film to the desired shape.
3	Heat-weld the LDPE films to form 2 cylindrical shaped tubes with 1 mm oversized compared with model, ie, 17-mm diameter cover for the 16-mm aortic diameter in the model and 9-mm diameter cover for the 8-mm iliac diameter in the model.
4	Suture the 2 parts to form the LDPE cover
5	Insert the collapsed stents into silicone tubes.
6	Insert collapsed stent into LDPE cover
7	Redeploy self-expanding stent suture ends of cover to stents after trimming.
8	Insert collapsed LDPE covered stent into silicone tubes.
9	Insert collapsed stent with LDPE cover into flow phantom
10	Redeploy the neo AFX self-expanding stent with LDPE cover in the flow phantom

Abbreviations: ePTFE, expanded polytetrafluoroethylene; LDPE, low-density polyethylene.

Appendix B

Author Contributions

Conception and design: Chong, Groot Jebbink, Jansen, and Sun. Analysis and interpretations: Chong, Groot Jebbink, and Mirgolbabaee. Data collection: Chong, Groot Jebbink, and Mirgolbabaee. Writing the article: Chong and Groot Jebbink. Critical revision of the article: Chong, Doyle, Groot Jebbink, Mirgolbabaee, Jansen, Reijnen, Sun, van de Velde, and Versluis. Statistical analysis: Chong. Overall responsibility: Chong, Doyle, Jansen, and Sun

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Michel Reijnen is a consultant with Endologix and Bentley Innomed.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Curtin University (Grant/Award Number: FieldTrip ST3483).

ORCID iDs

Albert Chong

Hadi Mirgolbabaee

Zhonghua Sun

Lennart van de Velde

Erik Groot Jebbink

Supplementary Material

The online videos are available at

References

Goverde

PCJM

Grimme

FAB

Verbruggen

PJEM

, et al. Covered endovascular reconstruction of aortic bifurcation (CERAB) technique: a new approach in treating extensive aortoiliac occlusive disease. J Cardiovasc Surg (Torino). 2013;54:383–387.

Reijnen

MM.

Update on covered endovascular reconstruction of the aortic bifurcation. Vascular. 2020;28:225–232.

Groot Jebbink

Mathai

Boersen

, et al. Hemodynamic comparison of stent configurations used for aortoiliac occlusive disease. J Vasc Surg. 2017;66:251–260.e1.

Chiu

KWH

Davies

RSM

Nightingale

, et al. Review of direct anatomical open surgical management of atherosclerotic aorto-iliac occlusive disease. Eur J Vasc Endovasc Surg. 2010;39:460–471.

Taeymans

Groot Jebbink

Holewijn

, et al. Three-year outcome of the covered endovascular reconstruction of the aortic bifurcation technique for aortoiliac occlusive disease. J Vasc Surg. 2018;67:1438–1447.

Lun

Zhang

, et al. Comparison of midterm outcomes between surgical treatment and endovascular reconstruction for chronic infrarenal aortoiliac occlusion. J Vasc Interv Radiol. 2015;26:196–204.

Maldonado

Westin

Jazaeri

, et al. Treatment of aortoiliac occlusive disease with the Endologix AFX unibody endograft. Eur J Vasc Endovasc Surg. 2016;52:64–74.

Sirignano

Mansour

Capoccia

, et al. Results of AFX unibody stent-graft implantation in patients with TASC D aortoiliac lesions and coexistent abdominal aortic aneurysms. J Endovasc Ther. 2017;24:846–851.

Van Haren

Goldstein

Velazquez

, et al. Endovascular treatment of TransAtlantic Inter-Society Consensus D aortoiliac occlusive disease using unibody bifurcated endografts. J Vasc Surg. 2017;65:398–405.

10.

Cury

MVM

Rodrigues

Issa

AAT

. Percutaneous implantation of a unibody endograft to treat non-aneurysmal aortoiliac disease. Vasa. 2017;46:138–141.

11.

Ford

Nikolov

Milner

, et al. PIV-measured versus CFD-predicted flow dynamics in anatomically realistic cerebral aneurysm models. J Biomech Eng. 2008;130:021015.

12.

Gallo

Gulan

Di Stefano

, et al., Analysis of thoracic aorta hemodynamics using 3D particle tracking velocimetry and computational fluid dynamics. J Biomech. 2014;47:3149–3155.

13.

Raschi

Mut

Byrne

, et al. CFD and PIV analysis of hemodynamics in a growing intracranial aneurysm. Int J Numer Method Biomed Eng. 2012;28:214–228.

14.

Buck

AKW

Groszek

Colvin

, et al. Combined in silico and in vitro approach predicts low wall shear stress regions in a hemofilter that correlate with thrombus formation in vivo. ASAIO J. 2018;64:211–217.

15.

Groot Jebbink

Grimme

FAB

Goverde

PCJM

, et al. Geometrical consequences of kissing stents and the covered endovascular reconstruction of the aortic bifurcation configuration in an in vitro model for endovascular reconstruction of aortic bifurcation. J Vasc Surg. 2015;61:1306–1311.

16.

Hughes

How

Pulsatile velocity distribution and wall shear rate measurement using pulsed doppler ultrasound. J. Biomech. 1994;27:103–110.

17.

Brindise

Busse

Vlachos

PP.

Density- and viscosity-matched Newtonian and non-Newtonian blood-analog solutions with PDMS refractive index. Exp Fluids. 2018;59:173.

18.

van de Velde

Donselaar

Groot Jebbink

, et al. Partial renal coverage in endovascular aneurysm repair causes unfavorable renal flow patterns in an infrarenal aneurysm model. J Vasc Surg. 2018;67:1585–1594.

19.

Suess

Anderson

Danielson

, et al., Examination of near-wall hemodynamic parameters in the renal bridging stent of various stent graft configurations for repairing visceral branched aortic aneurysms. J Vasc Surg. 2016;64:788–796.

20.

Dhawan

Nanjundappa

ARP

Branch

, et al., Shear stress and plaque development. Exp Rev Cardiovasc Ther. 2010;8:545–556.

21.

Allon

A patient with recurrent arteriovenous graft thrombosis. Clin J Am Soc Nephrol. 2015;10:2255–2262.

22.

Zarins

Giddens

Bharadvaj

, et al. Carotid bifurcation atherosclerosis. Quantitative correlation of plaque localization with flow velocity profiles and wall shear stress. Circ Res. 1983;53:502–514.

23.

Sotelo

Urbina

Valverde

, et al. 3D quantification of wall shear stress and oscillatory shear index using a finite-element method in 3D CINE PC-MRI data of the thoracic aorta. IEEE Trans Med Imag. 2016;35:1475–1487.

24.

De Keulenaer

Chappell

Ishizaka

, et al. Oscillatory and steady laminar shear stress differentially affect human endothelial redox state: role of a superoxide-producing NADH oxidase. Circ Res. 1998;82:1094–1101.

25.

Rayz

Boussel

, et al., Flow residence time and regions of intraluminal thrombus deposition in intracranial aneurysms. Ann Biomed Eng. 2010;38:3058–3069.

26.

Boersen

Groot Jebbink

Versluis

, et al. Flow and wall shear stress characterization after endovascular aneurysm repair and endovascular aneurysm sealing in an infrarenal aneurysm model. J Vasc Surg. 2017;66:1844–1853.

27.

Chong

Sun

van de Velde

, et al. A novel roller pump for physiological flow. Artif Organs. 2020;44:818–826.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB