Chapter 96
Nonaortic Stents and Stent-Grafts
Benjamin Pearce, William Jordan
Endovascular therapy has changed the landscape of vascular surgical practice. A major thrust into catheter-based therapy was initiated in 1999 with U.S. Food and Drug Administration (FDA) approval of endografts for aneurysmorrhaphy. As this important technique became ubiquitous, surgeons began to utilize stents in other vascular beds both for initial therapy and as secondary treatment after failed vascular grafts. Endovascular therapy with stents is now utilized throughout the vascular system, including arteries and veins ranging from the intracranial circulation to the tibial arteries.
As the use of endovascular stents has undergone more critical study, it has become evident that various vascular beds react differently to stent placement.1–3 Furthermore, angioplasty and stenting alter the biology of the treated vessel, a finding that has implications for both short-term and long-term results. The nature of the lesion, as well as the vascular bed to be treated, dictate specific nuances that guide the surgeon in choosing an appropriate stent.
This chapter focuses on factors that relate to the choice of stents and stent-grafts for endovascular intervention in nonaortic vascular beds. The indications for stent use in conjunction with balloon angioplasty as well as the interaction between vessel and stent are reviewed. The multiple factors that influence the optimal stent choice in a given circumstance are considered, including device characteristics relating to the mechanics of stent deployment, cell design, precision, treatment length, deliverability, and adjunctive stent designs that may affect therapy.
Historical Background
Isolated dilatation of vessels using dilatation catheters was first introduced in 1964 by Dotter et al.4 This technique was refined in the 1970s by Gruentzig et al,5 who used smaller catheters with attached balloons that could be delivered through the vascular tree from a remote location. Balloon angioplasty became a popular technique in the 1980s but remained inferior to surgical reconstruction because of the high acute occlusion rate and intermediate restenosis rate. Acute technical failure occurred as a result of elastic recoil, vasospasm, plaque rupture, or dissection. Recurrent stenosis, caused by an intense hyperplastic response, was commonly seen within the first 2 years after intervention. In initial series of coronary angioplasty, interval restenosis occurred in 30% to 50% of treated lesions.6–8 Likewise, angiographic failure of isolated angioplasty in the renal, iliac, and femoropopliteal arteries occurred in up to 26%,9,10 32%,1 and 50%,11,12 of cases, respectively.
Stenting was introduced to improve results of angioplasty by ensuring an adequate lumen, maintaining flow, and reducing embolic load. Achieving optimal luminal diameter lessens the impact of in-stent neointimal formation. However, paradoxically, the stent itself has inherent properties that alter the development of normal vascular intimal formation and can lead to maladaptive remodeling through direct intimal damage related to the procedure and interaction between the arterial wall and the stent itself.
Stent-Vessel Interaction
Vessel Injury
The degree of intimal response to stent placement has been linked directly to the extent of vessel injury. Sullivan et al13 used an experimental stent with beveled struts to demonstrate this negative remodeling effect in vivo, utilizing a swine model. This experimental stent was designed to violate the internal elastic lamina. In comparison with Palmaz stents deployed in control animals, the experimental stent was associated with significantly greater neointimal formation. Furthermore, the extent of vessel injury demonstrated a linear effect on the absolute neointimal formation.
In addition, an inflammatory response to stent placement has been demonstrated histologically. In an early autopsy series, Farb et al14 demonstrated inflammatory cell infiltration in the area of the vessel directly adjacent to the stent struts. The absolute number of inflammatory cells present was significantly increased when stent struts violated the internal lamina and penetrated into the lipid core of the plaque. Subsequent analysis demonstrated that inflammatory mediators were present more than 6 months after implantation. Another study showed that in, addition to the effect of the stent itself, bacterial contaminants introduced to the vessel wall during stent delivery also play a role in neointimal formation.15
Fluid Dynamics
Endovascular stents also alter the fluid dynamics of the stented segment. The most widely studied impact of stent placement on flow and negative remodeling is the creation of areas of low (<5 dyne/cm2) wall shear stress (WSS). Alteration in WSS occurs both as a result of a change in luminal diameter of the treated vessel and secondary to the presence of the stent itself. In a computational flow model, LaDisa et al16,17 demonstrated several characteristics of stent placement that create a low shear stress environment. The factor that led to the most significant increase in the proportion of vessel wall exposed to a low shear environment was overdistention of the stented segment. Although some degree of stent oversizing is necessary for appropriate apposition of the stent to the vessel wall, LaDisa’s group demonstrated a 13-fold increase in the total native vessel exposed to low WSS with 20% stent oversizing over that with 10% oversizing. The low WSS can subsequently lead to greater neointimal hyperplasia and recurrent stenosis. This finding is validated clinically because most surgeons aim for a 10% oversizing when stenting infrainguinal occlusive lesions.
Strut Characteristics
The tolerances for stent construction are very strict, and minimal alteration of the strut height can have a significant impact on shear stress and neointimal formation. Intuitively, the area of the vessel wall adjacent to the stent struts has the greatest potential for negative remodeling. Eddy currents created as blood flows over the stent struts lead to regions of low shear. This effect results in a proportional change in shear stress with alterations of strut thickness. Formation of neointima in such areas of low shear stress has been reproduced in several models, and the thickness of the resultant neointima correlates with strut coverage, configuration, and thickness.18–20 Sprague et al21 demonstrated that positive remodeling, as measured by endothelial cell migration, is hampered in low WSS environments. In normal shear models, migration of endothelial cells increased 2.5-fold within 1 week after implantation of steel struts onto the endothelial surface. However, in stented models with low shear, this migration was delayed up to several months.22 The clinical effect of these findings was demonstrated by the angiographic restenosis rates of two nearly identical coronary stents that differed only by stent heights of 50 µm and 140 µm. The stent with the lower-profile design had less restenosis.23
Stent Composition
The material composition of the stent also plays a role in neointimal formation. The most common bare metal stent components are stainless steel, nitinol (nickel and titanium alloy), cobalt chromium alloy, and tantalum. Although the actual mechanism of vessel injury from stent components is unclear, corrosive products from alloys have been found within sections of vessel wall. In addition hypersensitivity of some patients to certain metals has been observed.24 Palmaz et al25 demonstrated that galvanic currents are created within stented arteries and lead to corrosion and subsequent vessel injury.
Stent Types and Characteristics
Each stent has intrinsic properties that determine whether or not it is suitable for any given lesion.25–31 On the basis of the biologic interaction between vessel and stent, an ideal stent would be easy to deliver through a small-caliber sheath, be readily visible on fluoroscopy, conform to the vessel upon deployment, prevent acute procedural failure, provide long-term resistance to negative remodeling, and be fracture-resistant.
Generally, stents are divided into two groups according to their construction and mode of deployment—self-expanding (SE) versus balloon-expandable (BE) (Table 96-1).
Table 96-1
Comparison of Balloon-Expandable and Self-Expanding Stents Based on Intrinsic Characteristics and Clinical Use
| Characteristic | Balloon Expandable | Self Expanding |
| Radial force | High | Low |
| Flexibility | Low | High |
| Requires delivery sheath | Yes | No |
| Radiopacity | High | Variable |
| Oversizing recommended | No | Yes |
| Treats lesions with variable diameter | No | Yes |
| Resistant to external compression/bending | No | Yes |
Important Characteristics
In a comparison of SE and BE stents, the major characteristics that determine suitability are radial force, flexibility, and precision of deployment. Both SE and BE stents can be covered with polytetrafluoroethylene (PTFE) or polyester and such stent-grafts combine the advantages of a stent with those of a graft.
Radial force is defined as the force required to produce a 50% reduction in the luminal diameter of the stent. The radial force of the stent maintains its apposition to the vessel wall and tacks down intimal flaps that may obstruct flow. It also provides the support to resist immediate vessel recoil and acute occlusion. This outward force may lead to a better technical result than that of balloon angioplasty alone. As the stent is incorporated into the vessel, the radial force resists deformation and negative remodeling to maintain luminal diameter over time.
Ultimately, radial force is a product of both stent design and composition. The original Palmaz stent has a stainless steel slotted/diamond design. The slotted configuration allows the stent to maintain a low profile for loading on the balloon. Once expanded, the slots become diamonds, resisting further conformational change and providing a high radial force. The Wallstent (Boston Scientific, Natick, Mass.) also has a diamond configuration, but is designed to change lengths with its diameter. Therefore, its radial force is related to both its design and the degree of endothelialization within the artery. As the stent becomes more securely anchored, it resists shortening and actually leads to increased radial force. Various balloon-expandable and self-expanding stents are shown in Figure 96-1.
Figure 96-1 Comparison of stents. From top to bottom are the balloon-expandable Palmaz stent, the self-expanding nitinol Symphony stent, and the self-expanding stainless steel Wallstent.
Conversely, nitinol stents rely on the inherent nature of their metallic composition to provide resistance to deformation. The nitinol alloy assumes a predetermined configuration at a desired temperature. At low temperatures, the alloy exists in the martensite state (metallurgic property of shape-memory alloys in which the crystalline structure of the alloy is elongated or asymmetric at cooler temperatures), which is flexible and aids both mounting on the catheter shaft and deliverability. At higher temperatures, the alloy adopts a crystalline austenite state (metallurgic property of shape memory alloys in which the crystalline structure is symmetric) which makes the stent rigid, thereby providing more radial force.
Flexibility is determined by the same properties that govern radial force. BE stents require a force to change conformation. Thus these stents are less able to maneuver through tortuous vessels. BE stents are susceptible to deformation in mobile arteries because the torque required to maneuver them may enact a conformational change in the stent. In contrast, the nitinol stent, while in the martensite state, can be easily deformed and is therefore more suitable for arterial anatomy, which requires significant maneuverability to reach the target lesion. Interestingly, nitinol may change states not only in the setting of temperature change, but also in response to external compression. Such qualities afford nitinol stents improved flexibility in mobile arteries after implantation. Such is not the case for BE stents (Fig. 96-2).
Figure 96-2 This balloon-expandable stent in the cervical portion of the carotid artery is deformed at the distal endpoint where the carotid artery is mobile and susceptible to external compression.
Radiopacity is a consideration important in stent deployment. The material used in stent construction is a major determinant of radiopacity. Stainless steel is more visible than nitinol; thus BE stents are, generally, more visible on radiographic imaging than SE stents. Improved visibility leading to increased accuracy of deployment must be considered in the choice of the appropriate stent for a given indication (Fig. 96-3).
Balloon-Expandable Stents
BE stents are optimal in clinical scenarios where high radial force and precise deployment are imperative. Unfortunately, the characteristics of BE stents may also limit deliverability to target lesions. Because they are mounted on balloons, BE stents are at risk for dislodgement both during transit to their target and while crossing the lesion. For this reason, it is recommended that a BE stent be delivered through a guiding catheter, or sheath, over a stiff guide wire. The guiding sheath should be advanced beyond the target lesion before the stent is delivered to the deployment site. Once the stent is positioned at the lesion, the sheath is retracted to allow for stent deployment. The BE stent is deployed as the balloon is inflated to nominal pressure. The balloon is designed to inflate at both ends simultaneously, so as to prevent a forward slippage (“watermelon seeding”) of the stent off the balloon before complete deployment.
The BE stent achieves a size corresponding to the degree to which the balloon is expanded. Because this stent is deployed by expansion, vigorous oversizing of the stent in relation to the vessel is not recommended. However, one advantage of the BE stent is that it can be further distended to a larger diameter if optimal vessel apposition is not obtained. Of note, aggressive oversizing of the stent with a larger balloon will foreshorten the stent and weaken its struts, leading to increased risk of stent failure.
Innate characteristics of BE stents, including minimal displacement with deployment, make them ideal for treating difficult lesions in which precise placement is paramount. Thus, BE stents are widely used to treat ostial lesions of aortic branch vessels, such as renal, common iliac, and subclavian arteries, where the lesions are anatomically fixed and thus have little potential for stent deformation with movement. In addition, plaque at bifurcations tend to be more calcific and prone to dissection. In our experience, deployment of “kissing” BE stents in such lesions results in a lower incidence of dissection than balloon angioplasty alone.
The construction of the BE stent also accounts for its major limitations. When the segment being treated transitions across a branch point from a larger into a smaller artery, the high radial force across arteries that have different radii make a BE stent inappropriate; examples of such branch points include the iliac and carotid bifurcations. In these settings, the rigid size constraints of the BE stent may not allow for adequate apposition to the entire treated segment. Furthermore, these transitions are frequently mobile and may lead to stent deformation. Additionally, the higher profile needed to deliver both stent and balloon may necessitate predilation for delivery of the stent to the desired location. In certain situations, such as treatment of renal and carotid stenoses, the consequences of embolization are severe. Predilation should be considered with small balloons to improve subsequent stent deliverability and to prevent embolism or dissection. Embolic protection devices can also be considered in territories at risk of embolic complications. Finally, a BE stent may fracture or retain a new conformation when stressed by outside forces. This risk of deformation affects the decision to use a BE stent across the inguinal ligament or joints. In these positions, the risk of stent compression with fracture and subsequent vessel occlusion is increased. Care must also be taken when one is operating on an artery treated with a BE stent. Clamping of such an artery may lead to a permanent conformational change or stent fracture that may require redilatation or surgical removal of the stent.
Self-Expanding Stents
Self-expanding stents are better suited for tortuous lesions or those traversing vessels of variable diameters. These stents are more flexible than BE stents (Fig. 96-4). Therefore, they can be delivered through vessels that create more torque within the catheter. To maintain the stent in its constrained form during transit to the lesion, the stent is covered by an outer sheath on the mounting catheter. When this sheath is withdrawn, the stent is allowed to take its natural shape. As such, SE stents do not require that guiding catheters or sheaths be advanced beyond the target lesion. Actual deployment of a SE stent requires sequential removal of the constraining sheath from the distal end of the catheter to its proximal end. During this maneuver, the stent can either inadvertently jump forward or be retracted by the operator. This potential jumping forward of the stent can lead to maldeployment and must be considered by the operator prior to final deployment maneuvers.
Figure 96-4 Plain radiographs of a stainless steel self-expanding stent (A) and nitinol self-expanding stent (B) show the improved visibility of the stainless steel stent.
Another factor to consider in the use of an SE stent is the need for adequate oversizing. As mentioned previously, flow dynamic models demonstrate optimal shear environments at a 10% oversizing. Nevertheless, it is essential to choose a stent that opposes the vessel wall on deployment because an SE stent cannot be overdilated. In the instance in which a self-expanding stent is undersized for a given vessel, a second, larger-diameter stent can be deployed within the initial stent to aid with stent fixation. A BE stent is the preferred choice in this circumstance because its increased radial force can overcome the nominal diameter of the SE stent.
Because SE stents have less radial force and, conversely, improved elasticity, they are ideal for lesions that cross tortuous vessels, joints, or vessel branches. Common locations where SE stents are used include the carotid bifurcation and the transition zone from the common to the external iliac artery. SE stents are used preferentially in infrainguinal arteries because they can accommodate movement within the limb with less likelihood of fracture and subsequent failure.
Cell Size: Open and Closed Cells
Stent cell size should be considered in the selection of the best stent for a given clinical situation. Cell size refers to the area outlined by connected metallic components within a stent. Stents with a large cell size are labeled “open cell” stents, and those with a small cell size are termed “closed cell” stents (see Figure 96-1). The cell size of a stent and the connection of the metallic wires to each other may influence performance.
A closed cell stent has consistent interconnection of all stent wires throughout its length. This consistency provides a fixed area of interstices within the stent and uniform coverage of the vessel wall. Such construction decreases the free cell area and, may trap fractured plaque at the time of deployment, limiting distal embolization. This configuration, however, makes a stent less flexible and conformable. Considering the previously mentioned advantage of the SE stent, the closed cell stent may have more difficulty matching the tortuosity and change in luminal diameter across a vessel length.
