Stent Graft Configurations Research Articles

Shear accumulation as a means for evaluating risk of thromboembolic events in novel endovascular stent graft designs

ObjectiveThis study proposes to establish a simulation-based technique for evaluating shear accumulation in stent grafts and to use the technique to assess the performance of a novel branched stent graft system. MethodsComputational fluid dynamics models, with transient boundary conditions, particle injection, and rigid walls, simplifying assumptions were developed and used to evaluate the shear accumulation in various stent graft configurations with a healthy aorta as comparison. ResultsShear streamlines are presented for the various configurations. Shear accumulation was also calculated for each configuration. The number of particles with shear accumulations >3.5 Pa-s for each configuration was compared with the shear accumulation values of commercially available mechanical aortic valves from the literature. ConclusionsThe stent graft configuration with the diaphragm does have particles with shear accumulation >3.5 Pa-s. However, the percentage of particles with shear accumulation above 3.5 Pa-s is less than the two commercially available mechanical aortic valves, and more surprisingly, is smaller than in the healthy aorta.

Thoracic endovascular aortic repair (TEVAR) in proximal (type A) aortic dissection: Ready for a broader application?

ObjectiveThoracic endovascular aortic repair (TEVAR) has demonstrated encouraging results and is gaining increasing acceptance as a treatment option for aortic aneurysms and dissections. Yet, its role in managing proximal aortic pathologies is unknown—this is important because in proximal (Stanford type A) aortic dissections, 10% to 30% are not accepted for surgery and 30% to 50% are technically amenable for TEVAR. We describe our case series of type A aortic dissections treated by using TEVAR. MethodsBetween year 2009 and 2016, 12 patients with acute, subacute, or chronic type A aortic dissection with the proximal entry tear located between the coronaries and brachiocephalic artery were treated with TEVAR at 3 centers. Various stent-graft configurations were used to seal the proximal entry tear in the ascending aorta under rapid pacing. ResultsA total of 12 patients (9 male, 3 female), mean age 81 ± 7 years, EuroSCORE II 9.1 ± 4.5, underwent TEVAR for the treatment of type A aortic dissection. Procedural success was achieved in 11 of 12 patients (91.7%). There was 1 minor stroke and 1 intraprocedural death. No additional deaths were reported at 30 days. At 36 months, there were 4 further deaths (all from nonaortic causes). The mean survival of these 4 deceased was 23 months (range 15-36 months). Follow-up computed tomography demonstrated favorable aortic remodeling. ConclusionsTEVAR is feasible and reveals promising early results in selected patients with type A aortic dissection who are poor candidates for surgical repair. The current iteration of stent-graft technology, however, needs to be adapted to features specific to the ascending aorta.

Standard off-the-shelf versus custom-made multibranched thoracoabdominal aortic stent grafts

The complex aortic branch anatomy in thoracoabdominal aortic aneurysms (TAAAs) and pararenal aorticaneurysms (PRAAs) presents a challenge for endovascular repair. The multibranched endovascular device hasdurable midterm results with use of a custom branch stent graft (CSG) configuration. The midterm results with use ofthe standard branch stent graft (SSG) configuration are unknown, but it has the advantage of off-the-shelf technology. The goal of this study was to compare the midterm outcomes of CSG and SSG multibranched endovascular devices. From July 2005 to September 2014, 133 patients underwent elective endovascular repair of TAAA and PRAA in a prospective trial. Beginning in December 2008, SSGs were used in those with suitable anatomy. Fifty patients (mean age, 71± 7years; 11 women [22%]) were treated using SSGs, and 83 patients (mean age, 74± 9years; 22 women [26.5%]) underwent repair using CSGs. The SSG and CSG groups were similar with regard to aneurysm size, aneurysm extent, and medical comorbidities, with the sole exception of lung disease, which was more common in the SSG group. All stent grafts were deployed as intended, with no conversions to open repair. Mean± standard deviation follow-up (days) was 694± 525 for the SSG group and 942± 764 for the CSG group (P= .045). There were no significant differences in aneurysm-related death, renal failure requiring dialysis, stroke, endoleak, visceral or renal branch occlusion, lower extremity weakness, or reintervention (P > .05 for each). The volume of contrast material was significantly lower in those with SSGs compared with CSGs (P= .016), but there were no significant differences in operative or fluoroscopy times. Time to treatment (days from consent to surgery) was significantly lower in SSG patients compared with CSG patients (P= .01). For patients with suitable anatomy, the use of SSGs for TAAA and PRAA repair results in significantly shorter wait times to surgery and is as safe, effective, and durable in the midterm compared with CSGs.

Examination of near-wall hemodynamic parameters in the renal bridging stent of various stent graft configurations for repairing visceral branched aortic aneurysms

This study examined the flow behavior of four stent graft configurations for endovascular repair of complex aneurysms of the descending aorta. Computational fluid dynamics models with transient boundary conditions and rigid wall simplifying assumptions were developed and used with four distinct geometries to compare various near-wall hemodynamic parameters. Graphic plots for time-averaged wall shear stress, oscillating shear index, and relative residence time were presented and compared among the four stent graft configurations of interest. Abrupt 90° and 180° changes in stent geometry (particularly in the side branches) cause a high momentum change and thus increased flow separation and mixing, which has significant implications in blood flow characteristics near the wall. By comparison, longer bridging stents provide more gradual changes in momentum, thus allowing blood flow to develop before reaching the target vessel.

Geometry and respiratory-induced deformation of abdominal branch vessels and stents after complex endovascular aneurysm repair

This study quantified the geometry and respiration-induced deformation of abdominal branch vessels and stents after fenestrated (F-) and snorkel (Sn-) endovascular aneurysm repair (EVAR). Twenty patients (80% male; mean age, 75.2 ± 7.4 years; mean aneurysm diameter, 6.2 ± 1.8 cm) underwent computed tomography angiography during inspiratory and expiratory breath hold protocols after F-EVAR (n = 11) or Sn-EVAR (n = 9). Centerlines for the aorta and visceral vessels were extracted from three-dimensional models. Branch angles were computed relative to the orthogonal plane at the branch ostia, and end-stent angles of the left renal artery (LRA) and right renal artery (RRA) were computed relative to the distal stent orientation. The radius of peak curvature was defined by the circumscribed circle at the highest curvature. Sn-renal branches were more downward-angled than F-renal branches (P < .04). At the distal ends of the RRA stents, Sn-RRAs were angled greater than F-RRAs (P < .03) and had a smaller radius of peak curvature (P < .03). With expiration, the end-stent angle of Sn-LRAs increased by 4° ± 4° (P < .02) and exhibited a significant reduction of radius of curvature (P < .04). The unstented celiac arteries were more downward-angled (P < .02, inspiration), with a smaller radius of curvature (P < .00001), than the unstented superior mesenteric arteries. With expiration, the celiac arteries angled upwards by 9° ± 9° (P < .0005), which was greater than the superior mesenteric arteries (P < .03). At a median postoperative follow-up of 12.6 months (range, 1.0-37.1 months), branch vessel patency was 100%, serum creatinine levels remained stable, and one reintervention was required for a type III endoleak at the main body-LRA stent interface. Sn-renals were angled more inferiorly at the branch and more angulated at the stent end than F-renals due to stent placement strategies. Sn-LRAs exhibited a significant change in end-stent angle and curvature during respiration, a finding that may compromise long-term durability for parallel stent graft configurations. Further investigation is warranted to better optimize anatomic, patient, and branch vessel stent selection between fenestrated and snorkel strategies and their relationship to long-term patency.

Early versus late experience in fenestrated endovascular repair for abdominal aortic aneurysm

The objective of this study was to evaluate operative results and 1-year outcomes in early vs late experience after fenestrated endovascular aortic repair. All patients treated in Malmö, Sweden, and in Lille, France, with fenestrated endovascular repair for abdominal aortic aneurysm were prospectively enrolled in a computerized database. Early experience was defined as the first 50 patients treated at each center. Data from early and late experience were retrospectively analyzed and compared for differences in operative results and 1-year outcomes. Early experience covered 4.7 years in Malmö and 4.5 years in Lille; late experience covered 5.6 years in Malmö and 3.7 years in Lille. A total of 288 patients were included. In the later phase, stent graft configuration was more complex because of increased number of fenestrations/scallops incorporated in the graft design (2.7 ± 0.8 vs 3.2 ± 0.7; P < .001). Despite this, volume of contrast material and radiation time decreased by 27% and 20%, respectively, whereas procedure time remained unchanged. At 1 year, a trend toward decreasing abdominal aortic aneurysm diameter was observed in the late group, but no differences were found in mortality, endoleaks, or target vessel patency between the groups. With increasing experience, fenestrated endovascular aneurysm repair design has become more complicated, with more visceral vessels targeted for better proximal seal, while operative risk still remains low. Simultaneously, radiation time and volume of contrast material have been reduced, with possible long-term benefits for the patient.

Endovascular Repair of Abdominal Aortic Aneurysms in the Presence of a Transplanted Kidney.

To present our experience performing endovascular repair of abdominal aortic aneurysms in kidney transplanted patients. A retrospective review of all patients who underwent endovascular aneurysm repair (EVAR) for abdominal aortic aneurysms (AAA) performed at our institution from 2007 to 2014. We identified all patients who had previously undergone a kidney transplant. Data collected included: comorbidities, preoperative imaging modalities, indication for surgery, stent graft configurations, pre- and postoperative renal function, perioperative complications, and survival rates. A total of 267 EVARs were performed. Six (2%) had a transplanted kidney. Mean age was 74 (range, 64-82) years; five were males. Mean time from transplantation to EVAR was 7.5 (range, 2-12) years. Five underwent preoperative planning with noncontrast modalities only. Devices used included bifurcated (n = 3), aortouniiliac (n = 2), and tube (n = 1) stent grafts. Technical success was achieved in all patients. None experienced deterioration in renal function. Median follow-up was 39 (range, 6-51) months. Four patients were alive at the time of the study. Two patients expired during the period of follow-up from unrelated causes. EVAR is an effective modality for the management of AAAs in the coexistence of a transplanted kidney. It can be performed with minimal morbidity and mortality without harming the transplanted kidney. Special consideration should be given to device configuration to minimize damage to the renal graft.

Stent graft visualization and planning tool for endovascular surgery using finite element analysis

A new approach to optimize stent graft selection for endovascular aortic repair is the use of finite element analysis. Once the finite element model is created and solved, a software module is needed to view the simulation results in the clinical work environment. A new tool for interpretation of simulation results, named Medical Postprocessor, that enables comparison of different stent graft configurations and products was designed, implemented and tested. Aortic endovascular stent graft ring forces and sealing states in the vessel landing zone of three different configurations were provided in a surgical planning software using the Medical Imaging Interaction Tool Kit (MITK) software system. For data interpretation, software modules for 2D and 3D presentations were implemented. Ten surgeons evaluated the software features of the Medical Postprocessor. These surgeons performed usability tests and answered questionnaires based on their experience with the system. The Medical Postprocessor visualization system enabled vascular surgeons to determine the configuration with the highest overall fixation force in 16+/-6 s, best proximal sealing in 56+/-24s and highest proximal fixation force in 38+/- s. The majority considered the multiformat data provided helpful and found the Medical Postprocessor to be an efficient decision support system for stent graft selection. The evaluation of the user interface results in an ISONORM-conform user interface (113.5 points). The Medical Postprocessor visualization software tool for analyzing stent graft properties was evaluated by vascular surgeons. The results show that the software can assist the interpretation of simulation results to optimize stent graft configuration and sizing.

Hemodynamic variations due to spiral blood flow through four patient‐specific bifurcated stent graft configurations for the treatment of abdominal aortic aneurysms

Endovascular repair is now a recognised procedure for treating abdominal aortic aneurysms. However, post-operative complications such as stent graft migration and thrombus may still occur. To assess these complications numerically, the correct input boundary conditions, which include the full human aorta with associated branching, should be included. Four patient-specific computed tomography scanned bifurcated stent grafts (SGs) were modelled and attached onto a full human aorta, which included the ascending, aortic arch and descending aortas. Two of the SG geometries had a twisted leg configuration, while the other two had conventional nontwisted leg configurations. Computational fluid dynamics was completed for both geometries and the hemodynamics assessed. The complexity of the flow patterns and secondary flows were influenced by the inclusion of the full human aorta at the SG proximal section. During the decelerating phase significant recirculations occurred along the main body of all SG configurations. The inclusion of the full human aorta did not impact the velocity contours within the distal legs and there was no difference in drag forces with the SG containing the full human aorta and those without. A twisted leg configuration further promoted a spiral flow formation along its distal legs.

Computational Fluid Dynamics Evaluation of the Cross-Limb Stent Graft Configuration for Endovascular Aneurysm Repair

The technique of crossing the limbs of bifurcated modular stent grafts for endovascular aneurysm repair (EVAR) is often employed in the face of splayed aortic bifurcations to facilitate cannulation and prevent device kinking. However, little has been reported about the implications of cross-limb EVAR, especially in comparison to conventional EVAR. Previous computational fluid dynamics studies of conventional EVAR grafts have mostly utilized simplified planar stent graft geometries. We herein examined the differences between conventional and cross-limb EVAR by comparing their hemodynamic flow fields (i.e., in the "direct" and "cross" configurations, respectively). We also added a "planar" configuration, which is commonly found in the literature, to identify how well this configuration compares to out-of-plane stent graft configurations from a hemodynamic perspective. A representative patient's cross-limb stent graft geometry was segmented using computed tomography imaging in Mimics software. The cross-limb graft geometry was used to build its direct and planar counterparts in SolidWorks. Physiologic velocity and mass flow boundary conditions and blood properties were implemented for steady-state and pulsatile transient simulations in ANSYS CFX. Displacement forces, wall shear stress (WSS), and oscillatory shear index (OSI) were all comparable between the direct and cross configurations, whereas the planar geometry yielded very different predictions of hemodynamics compared to the out-of-plane stent graft configurations, particularly for displacement forces. This single-patient study suggests that the short-term hemodynamics involved in crossing the limbs is as safe as conventional EVAR. Higher helicity and improved WSS distribution of the cross-limb configuration suggest improved flow-related thrombosis resistance in the short term. However, there may be long-term fatigue implications to stent graft use in the cross configuration when compared to the direct configuration.

Endovascular repair of thoracoabdominal aneurysms: results of the first 48 cases.

In 2006, we began our experience with a novel technology for fully endovascular thoracoabdominal aneurysm repair, based on a custom-made, branched stent graft design. After 48 cases, we have learned and achieved substantial progress both in technical and in clinical skills. This paper describes the partial results of this ongoing experience. Patients in this series were selected for the presence of thoracoabdominal aortic aneurysms, with or without dissection, which was present in one patient. The observation of extensive anatomical variations in several patients prompted changes in many of the basic stent graft configurations, which are also described. Between August 2006 and June 2012, 48 patients were treated consecutively with custom-made branch stent grafts. The five patients with the longest follow-up available so far are at 71, 65, 60, 54 and 51 months post-procedure. The operative mortality rate, defined as death during or within a month of surgical hospitalization, was 21% (10 patients); each case is described herein. During postoperative follow up, nine patients died from causes not directly related to aneurysmal disease, at 3, 18, 20, 22, 24, 24, 37, 44 and 46 months. The main causes of death were myocardial infarction (four cases), cancer (two cases), gastrointestinal hemorrhage (one case), ischemic stroke (one case), and sepsis (one case). Permanent paraplegia occurred in one patient. It is still too soon to compare the results of endovascular repair of thoracoabdominal aneurysms with those of open surgical series. Despite the active and rapid progress currently observed for the endovascular method, it is still far from reaching its state-of-the-art plateau or becoming a gold standard. Further technological and technical advances in endovascular stent grafting seem to have a clear potential to provide very satisfactory operative outcomes for thoracoabdominal aortic aneurysms.

Determination of accurate stent graft configuration in abdominal aortic aneurysm using computed tomography: a preliminary study

An aortic stent graft is frequently used to cure an abdominal aortic aneurysm (AAA). It is critical to accurately fit the size and shape of the stent graft to the target region on the aorta. Proper sizing and shaping require the measurement of the orthogonal diameter of the target region from medical images. The present study aimed to acquire an accurate three-dimensional (3D) reconstruction of the aorta to determine the shape of the cross-sectional area where the stent graft would be implanted. A conventional geometric-active contour model was enhanced to prevent blurring and to improve edge detection with high noise resistivity. After the segmentation of two-dimensional (2D) images using the model, a 3D-reconstructed configuration of the aorta was achieved using a surface-rendering technique. The model could segment several selected synthetic images more accurately than conventional methods. Also, a 3D-reconstructed configuration of the abdominal aorta could be achieved using boundary coordinates extracted from 2D image segmentation. This preliminary study indicates the utility of the approach in optimizing stent graft configuration for AAA patients, thus enhancing stent graft healing.

In Situ Bending of a Thoracic Stent-Graft:A Proposed Novel Technique to Improve Thoracic Endograft Seal

To demonstrate the feasibility of a novel technique that modifies the configuration of a thoracic stent-graft after deployment to comply with the arch curvature. The principle of a Bowden cable has been applied to direct a conventional thoracic stent-graft in situ after deployment. A suture placed at the proximal inner curve of a conventional thoracic stent-graft is fitted with a sliding, self-locking knot attached to a line that runs inside a catheter through the central rod of the stent-graft. Traction applied to this line directs the endograft post deployment, which allows for better apposition to the aortic wall. Shortening the inner curve makes the stent-graft bend. The extent of bending is fully controlled by the surgeon and held in place with the sliding knot. A release mechanism allows removal of all luminal components of the mechanism. The described technique of directing a thoracic stent-graft in situ seems feasible and enables better apposition of the stent-graft in a glass model. It may improve the durability of thoracic stent-grafts in the aortic arch.

Transcarotid Artery Endovascular Reconstruction of the Aortic Arch by Modified Bifurcated Stent Graft for Stanford Type A Dissection

A 40-year-old man with Stanford type B dissection underwent his first endovascular repair (EVAR) in April 2004 by Talent thoracic stent graft. He had an uncomplicated recovery and maintained good blood pressure control. However, a new retrograde dissection appeared in September 2004. The new dissection involved his aortic arch and ascending thoracic aorta to the opening of the coronary arteries. To reconstruct the aortic arch, bypasses between the right common carotid artery (RCCA), left common carotid artery and left subclavian artery were performed before endovascular repair. A modified bifurcated Talent stent graft was deployed from the RCCA to the ascending thoracic aorta with a long limb in the innominate artery and a short limb in the aortic arch. A further two pieces of graft were deployed via the common femoral artery. The ascending thoracic aorta and aortic arch were reconstructed completely by the bifurcated stent graft. The final angiography confirmed that there was good stent graft configuration, normal blood flow, and stable haemodynamics. No endoleak or other major complications were encountered. This result indicated that it is possible to reconstruct the aortic arch with a bifurcated stent graft and could be a new endovascular repair model for complex thoracic aortic aneurysm and dissection.

The influence of aortic cuffs and iliac limb extensions on the outcome of endovascular abdominal aortic aneurysm repair

In a proportion of patients with an endovascular abdominal aortic aneurysm repair (EVAR), aortic cuffs or iliac graft limb extensions are required to enhance sealing or to fix the position of the device. This requirement arises when these goals are not primarily obtained with the basic stent-graft configuration. The aim of this study was to assess the influence of the use of endograft extensions during the primary EVAR procedure on the short- and long-term outcome. The study was based on the data of the EUROSTAR registry. Patient and anatomic characteristics, data regarding the procedure, postoperative complications, and the mortality of patients undergoing EVAR were retrieved from the database. Patients were divided into three groups: (1) no extensions, (2) proximal aortic cuffs, and (3) iliac limb extensions. Logistic regression and Cox proportional hazards models were used to compare significant influences of the use of cuffs or extensions on different outcomes relative to control patients, adjusted for patient and anatomic factors. The overall cohort comprised 6668 patients: 4932 (74.0%) without extensions, 259 (3.9%) with an aortic cuff, and 1477 (22.2%) with an iliac endograft extension. Both the 30-day (2.3%-3.9%) and the all-cause mortality rates (23%-27% at 4 years) were similar in the three study groups. The use of proximal cuffs or iliac extensions did not have an effect on the incidence of endoleaks of any type (24%-32% at 4 years). The incidences of device kinking (P = .0344) and secondary transfemoral interventions (P = .0053) during follow-up were increased in patients in whom iliac limb extensions were used. In patients with aortic cuffs, no significant associations with altered outcome were observed. The use of iliac graft limb extensions at EVAR was associated with a higher incidence of kinking and secondary transfemoral interventions, whereas proximal aortic cuffs did not influence outcome.

Endovascular stent-graft repair of ruptured aortic aneurysms.

To demonstrate the endovascular approach to the management of ruptured abdominal aortic aneurysms (AAA). From 1995 to 2001, 24 patients (21 men; mean age 69 years, range 26-92) underwent emergency endovascular treatment for ruptured AAA. The average interval between onset of symptoms and admission to the hospital was 8.0 hours; the mean time between admission and the operation was 2.3 hours. No suprarenal occluding catheter was used. The stent-graft configurations were 19 bifurcated, 4 tube, and 1 aortomonoiliac. Stent-graft placement was successful in 23 (96%) cases. Failed limb extension deployment prompted conversion to open surgery in the remaining patient. One case was converted to open surgery. Mean duration of treatment was 122 minutes. Three (12.5%) patients died in-hospital. The median hospital stay was 12 days. The rate of endoleaks (all type I) was 16.7%. The overall technical success rate was 77%. The 3-year actuarial survival rate was 75%. Our experience shows excellent results in emergency patients with ruptured AAAs treated with endovascular surgery. In order to verify these promising results, a broader-scale clinical study must be conducted.

Midterm outcomes of thoracic aortic stent-grafts: complications and imaging techniques.

To evaluate the midterm outcomes of thoracic aortic stent-grafting and the performance of computed tomographic angiography (CTA), radiography, and magnetic resonance angiography (MRA) in endograft surveillance. Forty-seven patients with traumatic thoracic aortic ruptures (n=16), aneurysms (n=14), false aneurysms (n=3), penetrating ulcers (n=3), and dissections (n=11) treated with stent-grafts were monitored in follow-up using chest radiography and CTA in all patients and MRA in 23 patients. Two perpendicular maximal aortic diameters, the sum of these diameters, and the elliptical cross-sectional area were determined and compared to baseline for the entire group and in subgroup analyses according to lesion type. CTA, MRA, and radiography were compared for their ability to detect endoleak, monitor stent-graft configuration, and measure aortic diameters. The mortality rate was 8.5%. Severe complications were observed in 14.8% (6% neurological complications); 12 (25.5%) patients had primary endoleaks. Over a mean 11-month follow-up (range 0.25-46 months), the aortic diameters decreased for all patients without endoleak (p<0.001). In the diameter/area subgroup analyses, only the traumatic rupture cohort demonstrated significant decreases in all 4 measurements. CTA and MRA measurements correlated well, but chest radiography was superior to both for visualizing stent-graft shape. In terms of endoleak detection, MRA missed only 1 (12.5%) endoleak (type II) seen on CTA; there were no false positive results with MRA. Morbidity and mortality observed after thoracic stent-grafting are acceptable. Radiography is better for monitoring stent-graft conformation, while CTA provides the best overall morphological information. The performance of MRA in endoleak detection is encouraging.

Changes in aneurysm morphology and stent-graft configuration after endovascular repair of aneurysms of the descending thoracic aorta

Objective: We sought to study changes in morphology and stent-graft configuration of descending thoracic aortic aneurysms after endovascular repair. Methods: Twenty-three patients treated with custom-made stent-grafts were studied. The stent-graft consisted of continuous, stainless-steel Z stents mounted within a polyester graft. In the last 11 cases the stents were interconnected with 3 longitudinal wires. Contrast-enhanced spiral computed tomography was performed preoperatively and at 1, 3, and every 6 months postoperatively. Angiography was used preoperatively and at 1-year follow-up. Proximal and distal necks were assessed for diameter and length. Aneurysm diameter, endoleaks, stent-graft migration, and changes in stent-graft configuration were evaluated. Results: During follow-up (median, 18 months; range, 1-48 months), excluded aneurysms decreased in diameter by 4 mm (0.5-10 mm, P =.0018). Endoleaks prevented size decrease. Five patients displayed neck dilatation, 4 at both the proximal and distal fixation sites and 1 only distally. In 7 (30%) patients there was proximal migration of the distal end of the stent-graft. Three (13%) patients displayed both distal migration of the proximal end of the stent-graft and proximal migration of the distal end of the stent-graft. There was a significant correlation between stent-graft kinking and appearance of proximal or distal stent-graft migration (P =.05 and P =.0007, respectively). In no case did the migration lead to appearance of an endoleak before intervention was performed. Conclusion: Excluded descending thoracic aortic aneurysms decrease in size on midterm follow-up. A subgroup of patients prone to neck dilatation might exist. A combination of neck dilatation and vector forces acting on stent-grafts in the tortuous thoracic aorta might lead to stent-graft migration.J Thorac Cardiovasc Surg 2001;122:47-52