Endovascular Repair of the Aorta: Stentgraft deformation matters
Jaimy Simmering is a PhD student in the department Multi-Modality Medical Imaging. (Co)Supervisors are prof.dr. R.H. Geelkerken and dr. E. Groot Jebbink from the faculty of Science & Technology and prof.dr. C.H. Slump from the faculty of Electrical Engineering, Mathematics and Computer Science.
Cardiovascular disease, which includes aneurysmal disease, remains the number two cause of death in the Netherlands. An aortic aneurysm (AA) is a local dilatation of the aorta of at least 1.5 times its original diameter. AAs most often develop inferior of the renal arteries in the abdominal aorta, i.e. as an abdominal aortic aneurysm (AAA). A rupture of an AA causes a large internal bleeding and is hence associated with mortality rates up to 80%. To prevent rupture, the AAs are treated by either open surgical repair (OSR) or endovascular aneurysm repair (EVAR). EVAR has become the gold standard for infrarenal AAA treatment according to the European Society for Vascular Surgery (ESVS), while the British National Institute for Health and Care Excellence (NICE) advices to consider EVAR only in select cases. This reluctancy of the NICE to set EVAR as the first choice in AAA management like the ESVS relates to the number of complications and reinterventions during follow-up that hamper treatment success. So is EVAR for infrarenal AAA associated with reintervention rates of 8-20% which even increases up to 12-44% for fenestrated EVAR (FEVAR) for pararenal AAA. For iliac branched devices (IBD) and thoracic EVAR (TEVAR) reintervention rates between 10-20% are reported. The reason for these high reintervention rates is stentgraft failure, expressed as endoleaks, graft migration, (partial) graft occlusion and ultimately AA rupture. To improve the outcomes of EVAR in general, we first need to improve our understanding of the behavior of the different stentgrafts. Several factors may influence the durability of these treatments, among which the cardiac cycle, the respiratory cycle, the body position and posture, blood flow and the composition of blood and the vascular wall. These factors may influence, or may be influenced by, stentgraft deformation. Stentgraft deformation can be defined as any change in shape, geometry and/or size of the stentgraft over time, for example radial diametral change (expansion) and bending (curvature change). The objective of this thesis was to broaden our understanding of the in situ behavior of different stentgraft platforms for different types of AAs in terms of stentgraft deformation during the cardiac cycle and during follow-up. Such deformation may be quantified using electrocardiogram (ECG)-gated computed tomography (CT) scans.
In Chapter 2, it was investigated to what extent the quantification of cardiac-pulsatility-induced displacement is comparable when the ECG-gated CT scans are reconstructed in 8 or 10 phases of the cardiac cycle. The quantification of cardiac-pulsatility-induced displacement was also investigated on two CT scanners from different manufacturers. The results revealed that the displacement quantification is comparable for ECG-gated CT scans reconstruction types and different scanners. The differences between scan reconstructions and scanner types remain below 0.3 mm, indicating that the displacement quantification is now hampered by the CT scan’s resolution.
In Chapter 3 we increased our understanding of the behavior of the limbs of a specific EVAR stentgraft: the Anaconda stentgraft system. The flexible, separate stent-ring design of the Anaconda limbs distinguishes it from other commonly used stent grafts. The design, in a way, mimics that of a vacuum cleaner hose with its individual circular stent-rings that provide it with flexibility while avoiding kinking. Deformation of the stentgraft limbs during the cardiac cycle and during follow-up was evaluated using data of a prospective observational study holding ECG-gated CT scans of 15 AAA patients. Throughout the follow-up period of two years, we observed an increase in limb curvature, shortening of the limbs, and a corresponding decrease in the distances between successive stent-rings. Meanwhile, the cardiac-pulsatility induced deformation remained fairly constant during follow-up with dynamic curvature changes ranging 1-2% of the mid-cardiac cycle value and pulsatile changes of inter-ring distances and limb length of ca. 0.3%. The flexible design of the Anaconda allows the stentgraft to be placed in patients with tortuous vascular anatomies that are not suitable for other EVAR devices. However, from this chapter the hypothesis arose that this design is likely to have the downside of inducing infolding of the graft material when the distances between the individual stent-rings significantly decrease due to increased curvature and/or reduced length of the limbs. Between these folds of the graft material, blood may become static and/or have turbulent flow, which may result in thrombus formation and thereby may contribute to the increased rate of limb occlusions observed for this particular type of stentgraft. This phenomenon of fabric folding/plication formation is also called the Concertina effect.
The hypothesis of the Concertina effect in the Anaconda limbs was further explored in Chapter 4 by evaluating a cohort of 84 patients with an AAA that was treated with an Anaconda stentgraft. The results showed that risk for thrombo-embolic events (limb occlusion and distal thromboembolisms originating from the stentgraft) after EVAR with the Anaconda stentgraft decreases for patients with more circumferential common iliac artery (CIA) calcification and a stronger decrease in curvature and tortuosity index after EVAR. These findings further support the hypothesis that flexibility of this stentgraft design may induce the Concertina effect that contributes to the development of thrombo-embolic events. Based on this research and with the intention to reduce the number of thrombo-embolic events, the instructions for use of the Anaconda were updated with the advice to place the stentgraft limbs in a stretched way and thereby avoid redundant fabric in the limb lumen at the risk of thrombus development.
The majority of the complications after FEVAR are either related to the bridging stents or to the proximal sealing and fixation of the main device. The deformation during the cardiac cycle and during follow-up of 19 fenestrated Anaconda stentgrafts with V12 bridging stents configurations in patients with a complex AAA was discussed in Chapter 5. Limited deformation in terms of end-stent angle, curvature and tortuosity index was observed, which indicates that these FEVAR configurations can be considered stable and durable. However, during follow-up the renal branch angles increased towards a perpendicular orientation to the aorta. This warrants for careful consideration of significantly increasing renal branch angles and other geometry alterations as this may promote the development of bridging stent related complications. These geometry changes are likely to be compensated more downstream in the target arteries with potential flow disturbances as a result.
In Chapter 6 the proximal sealing ring deformation during the cardiac cycle and follow-up of both the infrarenal and fenestrated Anaconda stentgrafts was quantified and compared. Limited pulsatile expansion was observed at all follow-up time points, which is considered beneficial for the device durability. However, the pulsatile displacement, expansion and bending of the proximal sealing rings was not uniform, but rather less on the posterior side of the aorta. This asymmetrical deformation should be taken into account in device manufacturing and selection, to ensure proper sealing and fixation of the stentgraft. Thereby, the methodology described in this chapter may also aid in identifying deviating local deformation patterns of the stent-rings that may point towards regions at risk for stent fatigue or pending endoleaks as a result of inadequate sealing due to, for example, local aortic wall weakening.
In Chapter 7 the cardiac-pulsatility-induced deformation of the aorto-iliac trajectory before and after placement of the Gore Excluder Iliac Branch Endoprosthesis (IBE) was evaluated. The IBE dampened the cardiac pressure wave along the investigated aorto-iliac trajectory. However, the displacement of in the internal iliac artery (IIA) appeared to have increased after IBE placement. The CIA-IIA trajectory was shortened and straightened postoperatively, which was observed as a reduction in curvature, length and tortuosity index. The decreased cardiac-pulsatility induced change in CIA-IIA curvature indicated stiffening of the trajectory. This stiffening may induce a compliance mismatch between the stented and unstented arteries, which could lead to local flow disturbances that may induce thrombus formation and other complications.
The postoperative behavior of the IBE was compared to that of a more stiff IBD, the Cook Zenith Bifurcated Iliac Side (ZBIS), in Chapter 8. This comparison yielded more pronounced pulsatile displacement during the cardiac cycle in the IBE than in the ZBIS throughout the CIA-IIA trajectory, especially in the cranio-caudal direction. Thereby, the IBE turned out to be more flexible and conformable than the ZBIS. This could be explained by the use of self-expandable stents, such as the dedicated IIA-component of the IBE, which are generally less stiff and thereby more conformable than balloon-expandable stents, such as the V12 stents that were used as ZBIS IIA-component.
In Chapter 9, a new terrain was explored by investigating the deformation of the aortic arch and its side branches after branched TEVAR (BTEVAR). The presented case was of a patient with an aortic arch aneurysm that was treated by BTEVAR and followed with ECG-gated CT scans. The pulsatile diametric expansion of the aortic arch decreased over time, which suggests that the aortic arch is stiffened. The increased stiffness of the aortic arch after BTEVAR reduces the Windkessel function of the aorta and increases the resistance for the blood to flow into the aorta. The continuing decrease of pulsatile diametric expansion during follow-up, may indicate that the cardiac function declines as a result of the increased stiffening and cardiac load. Thereby, in this patient an increased mismatch of pulsatile displacement at the end of branch stent was observed after BTEVAR. The ends of the branch stents showed significantly more displacement than the same locations preoperatively and the first bifurcations of the branch arteries. This mismatch may lead to either the branch stents scraping along the vessel wall or “pulling” the vessel wall each heartbeat. Both these consequences might lead to increased pulsatile wall stress and consequent intima injuries that produce micro-embolisms that may explain the ongoing cerebral infarction after BTEVAR. Therefore, careful patient and conformable (branch) stentgraft selection is of critical importance in BTEVAR.
Aside from the stentgraft deformation during the cardiac cycle and follow-up, other factors may contribute to the outcomes of EVAR procedures. One of these factors is the position and posture of the human body. The human body is not constantly in extended supine position as it is during CT scanning, but it constantly assumes different positions and postures during the day which may affect position and shape of the aorta and the stentgraft. The first step in investigating the extent of postural influence may be to visualize the aorta in different body positions and postures. However, the scoping review presented in Chapter 10 of this thesis revealed that this has only been done limitedly. Especially with the increasing number of endovascular treatments of arterial pathologies it would be of interest to quantify the deformation of the target vessels in different body positions so this can be taken into account in treatment (planning) and stent manufacturing.
In conclusion, this thesis showed the potential relation between clinical outcome and stentgraft deformation using detailed analysis of ECG-gated CT scans. The motion patterns and geometrical variables were calculated for different types of EVAR throughout the entire aorta. These findings can be translated to clinical conduct and stentgraft design and development. So is it now recommended to place Anaconda limbs in a stretched manner to prevent the Concertina effect, as described in the instructions for use, and is the stiffness of branch/bridging stents for FEVAR, IBD and BTEVAR a point of discussion and improvement for both the treating physicians and the stentgraft manufacturers. Still, stentgrafts produced for the same purpose may demonstrate different in situ behavior due to differences in stentgraft design and patient anatomy. This emphasizes the relevance of in vivo deformation analysis of the aorta and stentgrafts, which is also paramount for accurate stress-strain analysis, stent fatigue evaluation and design verification and should therefore be part of Conformité Européenne (CE, European Conformity) and Food and Drug Administration (FDA) approval procedures. Still, to fully comprehend the success and failure of EVAR stentgrafts, other contributing factors should be taken into account as well, such as the respiratory cycle, body position and posture, and physiological and biochemical aspects of the blood and arterial wall.