The concept of vascular stenting originated with Charles Dotter in 1969 but didn't become a clinical reality until the late 1980s. Nearly 30 years have passed since the Dotter concept, yet it has only been during the past decade that we have started to understand those factors governing stent-related patency and restenosis. The following review summarizes what we know about vascular stents, as well as some of the new approaches that may improve the patency of future devices.

What stents do and don't do

Stents are really just braces that prevent elastic recoil of tissue after angioplasty. Stents hold an angioplasty site open, tack back a dissection flap, and counteract the compressive effects of extrinsic masses and fibrosis. This is illustrated in figure 1, showing poor results following iliac angioplasty that have been improved by placement of a Palmaz stent. Beyond iliac artery intervention, stents have been used in a number of other vascular situations to increase procedural success. For instance, the technical success of ostial renal vein angioplasty is quite variable, but most interventional radiologists would agree that it ranges from 30 to 70%. This can be increased to greater than 95% with the addition of a Palmaz stent.1 On the venous side, subclavian vein angioplasty is approximately 75% successful in achieving a less than 50% rate of residual stenosis.2 The procedural success is increased to greater than 95% with the addition of a stent.3-5

While stents may dramatically improve the immediate post-angioplasty appearance in a number of locations, they are not able to prevent restenosis. A stent is no more than mesh scaffolding through which tissue may propagate. The process of restenosis, in fact, starts when a stent is placed, due to vascular "injury" that is induced by PTA and/or stent placement followed by deposition of fibrin, platelets, and leukocytes on the metallic stent struts.6-8 The processes which contribute to restenosis are complex and beyond the scope of this review. Nevertheless, it is fair to state that if vessel wall trauma does not incite an overly exuberant neointimal response, and if the buildup of fibrin and thrombus is not excessive, most stents will be incorporated with a thin, non-stenotic neointimal layer. In humans, this neointimal tissue consists of a collagen matrix, smooth muscle-like cells, and a luminal layer of endothelial cells.6,7 On the other hand, exuberant neointimal tissue formation may lead to stenosis.

Factors that lead to neointimal stenosis

When the neointimal response is exuberant, stenosis may occur within a stent. This is called stent-related restenosis or in-stent restenosis (figure 2). There are a number of factors that portend to stent-related restenosis. "Excessive" trauma at the time of angioplasty and/or stent placement is a proven method for inducing neointimal stenosis in a number of animal models, including the rat carotid artery and the pig coronary and iliac circulation. Care should be taken not to overdilate the initial stenotic lesion or the adjacent "normal" segments of vessel. Additionally, stents should be placed with minimal trauma to the angioplasty site, and overexpansion of the stent should be avoided. In our lab, the characteristic finding following stent placement in the canine iliac artery has been disruption of the internal elastic lamina at the ends of stented segment-not at the middle of the stent. We believe that this represents trauma induced by balloon dilation of the stent at the time of deployment (for balloon expandable stents) or after deployment of self-expanding stents when the stent is apposed to the vessel wall. We also have noted thicker neointima at the ends of these stents compared to the middle of the stents. We believe this is, in part, due to vessel wall trauma at the time of stent placement.

Another factor related to in-stent stenosis is adherence of fibrin and thrombus to the stent struts. Excessive thrombus can serve as the substrate for exuberant neointima. Smooth muscle cells may migrate into maturing thrombus, ultimately leading to an area of neointimal stenosis. Unfortunately, angiography is an insensitive way to look at deposition of thrombus and cells on a stent at the time of placement, and there is no way to assess thrombus deposition after the procedure has been completed. Intravascular ultrasound is far more sensitive for angiographically occult thrombus but is rarely used in a systematic manner for this purpose.

Some patients may be more susceptible to accelerated atherosclerosis and/or neointimal hyperplasia at the stented site. Factors that cause any given patient to be at risk remain to be defined, but it has often been observed that patients who continue to smoke and/or have diffusely small arteries seem to have an increased risk of restenosis involving stented peripheral arteries. Some patients with in-stent restenosis will ultimately be found to have metabolic abnormalities, such as primary and secondary hyperlipidemic states, hypercoagulable states (such as antiphospholipid-antibody syndrome), and disorders of the fibrinolytic system (including elevated lipoprotein(a) and homocystinuria).

Finally, a number of lesion types and locations are particularly prone to stent-related restenosis. This is certainly true for renal failure patients with AV shunts who have subclavian vein stenosis. In one recent series, the 6-month patency for subclavian vein stents was 42%, and at one year only 25% of the stented central veins were not stenosed or occluded.5 These disappointing results were found despite 100% technical stenting success with 10 and 12 mm stents. Clearly, the subclavian location is one of the sites where stent patency is reliably poor, perhaps due to continuous mechanical stress from extrinsic muscle and bony impingement upon the stented vein. Patency rates for stented AV shunt stenoses (figure 3) are only slightly better, with a reported 12-month patency of 20 to 40%.3,4,9,10

On the arterial side, femoropopliteal stenting has had poor patency in most series. Gray et al reported that only 22% of the femoropopliteal Wallstents placed in 58 limbs maintained primary patency at a year.11 Poor durability of stents in this circulation may have been related to a number of factors, including the long segments of diseased vessels that were stented. Nevertheless, most interventionalists are disappointed with the long-term patency of femoropopliteal stenting.

Transjugular intrahepatic portosystemic shunts (TIPS) are prone to restenosis, with a one-year primary patency of 25 to 66%.12,13 There are 2 predominant sites of TIPS stenosis: one at the hepatic vein end of the stent and the other in the mid-stent (figure 4). Most cases of stent-related TIPS stenosis can be treated simply by repeat angioplasty and sometimes with placement of an additional stent. However, a small subset of patients with TIPS-related stenosis will develop recurrent stenosis in their TIPS shunt despite angioplasty and restenting.

No current stent design favors patency

It would be exciting to find stent-related factors that could be modified to significantly lower the risk of restenosis within the stented segment. Theoretically, a number of designs and materials offer advantages that should improve stent durability and limit in-stent restenosis. Unfortunately, the following variables make little difference in animal models and, in some cases, human clinical trials:

1) flexible vs rigid stents14

2) self-expanding vs balloon-expanded stents14

3) stent metal (nitinol, Elgiloy, steel)14

4) articulated vs non-articulated stents15

5) degree of radial force16

6) amount of metal surface area17

7) heparin coating18

Of course, there are some limitations to the list. No one would argue that a long solid lead tube stent would perform as well as a flexible mesh nitinol stent. Within the range of currently available stent designs and materials, however, there is no proven difference in stent-related restenosis or occlusion. In addition, neither aggressive systemic 4-week

anticoagulation (coumadin with initial heparin, and aspirin) nor antiplatelet regimens (aspirin and ticlopidine) improved 6-month angiographic patency in a prospective coronary artery stent trial.19

Currently, our choice of stents for vascular use is based upon appropriateness of the stent for a specific lesion (for example, diameter, length and, perhaps, flexibility), ease of stent delivery, cost, and (often) whether the stent has been approved for the given indication. In my view, bare metal stents are an important but interim step in the development of devices for achieving durable percutaneous revascularization.

Possible approaches to limit neointimal in-stent restenosis

A number of strategies to limit stent-related restenosis are under evaluation. For medium size vessels, 5 to 8 mm in diameter, covered stents or "stent-grafts" may provide a physical barrier to neointimal ingrowth (figure 6). A variety of synthetic graft materials are being explored, including polyethylene terephthalate (PET), polyurethanes, and expanded polytetrafluoroethylene (ePTFE). Both PET and ePTFE have been used extensively as surgical bypass conduits, but their use as endoluminal prostheses for occlusive disease is only recently being explored. We have investigated ePTFE and PET stent-grafts in a canine model,20,21 and have noted a number of features related to healing and patency associated with these different devices. There are currently 2 stent-graft trials for occlusive disease in humans: one using a PET stent-graft (Wallgraft, Schneider, Minneapolis, MN) and the other testing an ePTFE stent-graft (HemoBahn, ProGraft, W.L. Gore and Associates, Flagstaff, AZ). A polyurethane stent-graft trial (Corvita Endoluminal Graft, Corvita, Miami, FL/Schneider) has been suspended by the sponsor pending modifications.

Another promising approach is vascular irradiation, exposing the stented segment to therapeutic levels of ionizing radiation. This can be accomplished with radioactive stents, catheter-delivered radiation, or external beam irradiation.22 A series of coronary trials are underway, but peripheral studies have been slow to begin.

A number of pharmacologic and immunologic approaches are being considered as stent adjuncts. The efficacy of local drug or antibody delivery to the stented segment is being evaluated with the hope of interfering with early fibrin/platelet formation on the stent (IIb/IIIa receptor inhibition) or smooth muscle cell proliferation and migration (for example, Taxol). If any of these approaches prove successful in limiting in-stent restenosis, there will no doubt be interest in preparing pharmacologically active stents.

Oral (systemic) medication would be an attractive alternative to local delivery if an agent can be found that limits restenosis. Cilostazol (an antithrombotic agent), Angiopeptin (an antiproliferative agent), and a variety of oral anti-IIIb/IIIa agents all are being studied. Is this a search for the "Holy Grail"? Perhaps, but maybe we truly are closer than ever to an oral medication that limits restenosis.

Regardless of how the technology ultimately develops, the next generation of vascular "stents" will offer improved patency by linking the intrinsic mechanical support of a stent with some other physiologic or mechanical inhibitor of neointimal proliferation. Exactly which new approach is most likely to succeed is not yet known. In fact, it is likely that several adjunctive approaches will prove valuable. Even if covered stents have no use in small vessels, combined brachytherapy and oral medication will limit neointimal formation. These are dynamic times, and the only certainty is that the contemporary interventionalist can no longer ignore the pharmacology, cell biology, and biomaterials data that is exploding in the vascular literature. AR

References

1. Rees CR, Palmaz JC, Becker GJ, et al: Palmaz stent in atherosclerotic stenoses involving the ostia of the renal arteries: Preliminary report of a multicenter study. Radiology 181:507-514, 1991.

2. Glanz S, Gordon DH, Lipkowitz GS, et al: Axillary and subclavian vein stenosis: Percutaneous angioplasty. Radiology 168:371-373, 1988.

3. Quinn SF, Schuman ES, Hall LMC, et al: Venous stenoses in patients who undergo hemodialysis: Treatment with self-expandable endovascular stents. Radiology 183:499-504, 1992.

4. Gray RJ, Horton KN, Dolmatch BL, et al: Use of Wallstents for hemodialysis access-related venous stenoses and occlusions untreatable with balloon angioplasty. Radiology 195:479-484, 1995.

5. Vesely TM, Hovsepian DM, Pilgram TK, et al: Upper extremity central venous obstruction in hemodialysis patients: Treatment with Wallstents. Radiology 204:343-348, 1997.

6. Schatz RA: A view of vascular stents. Circulation 79(2):445-457, 1989.

7. Palmaz JC: Intravascular stents: Tissue-stent interactions and design considerations. AJR 160:613-618, 1993.

8. Parsson H, Cwikiel W, Johansson K, et al: Deposition of platelets and neutrophils in porcine iliac arteries after angioplasty and Wallstent placement compared with angioplasty alone. Cardiovasc Interv Radiol 17:190-196, 1994.

9. Antonucci F, Salomonowitz E, Stuckmann G, et al: Placement of venous stents: Clinical experience with a self-expanding prosthesis. Radiology 183:493-497, 1992.

10. Vorwerk D, Guenther RW, Mann H, et al: Venous stenosis and occlusion in hemodialysis shunts: Follow-up results of stent placement in 65 patients. Radiology 195:140-146, 1995.

11. Gray BH, Sullivan TM, Childs MB, et al: High incidence of restenosis/reocclusion of stents in the percutaneous treatment of long-segment superficial femoral artery disease after suboptimal angioplasty. J Vasc Surg 25:74-83, 1997.

12. LaBerge JM, Ping EJ, Gordon RL, et al: Creation of transjugular intrahepatic portosystemic shunts with the Wallstent endoprosthesis: Results in 100 patients. Radiology 187:413-420, 1993.

13. Lind CD, Malisch TW, Chong WK, et al: Incidence of shunt occlusion or stenosis following transjugular intrahepatic portosystemic shunt placement. Gastroenterology 106:1277-1283, 1994.

14. Schurmann K, Vorwerk D, Kulisch A, et al: Neointimal hyperplasia in low-profile nitinol stents, Palmaz stents, and Wallstents: A comparative experimental study. Cardiovasc Interv Radiol 19:248-254, 1996.

15. Mundra H, Regar E, Klauss V, et al: Serial follow-up after optimized ultrasound-guided deployment of Palmaz-Schatz stents. Circulation 95:363-370, 1997.

16. Vorwerk D, Redha F, Neuerburg J, et al: Neointima formation following arterial placement of self-expanding stents of different radial force: Experimental results. Cardiovasc Interv Radiol 17:27-32, 1994.

17. Newman VS, Berry JL, Routh WD, et al: Effects of vascular stent surface area and hemodynamics on intimal thickening. J Vasc Interv Radiol 7:387-393, 1996.

18. Serruys PW, Emanuelsson H, van der Giessen W, et al: Heparin-coated Palmaz-Schatz stents in human coronary arteries: Early outcome of the Benestent-II Pilot Study. Circulation 93:412-422, 1996.

19. Kastrati A, Schuhlen H, Hausleiter J, et al: Restenosis after coronary stent placement

and randomization to a four-week combined antiplatelet or anticoagulant therapy: Six-month angiographic follow-up of the intracoronary stenting and antithrombotic regimen (ISAR) trial. Circulation 96:462-467, 1997.

20. Dolmatch BL, Tio FO, Li XD, Dong YH: Patency and tissue response related to two types of polytetrafluoroethylene-covered stents in the dog. J Vasc Interv Radiol 7:641-649, 1996.

21. Dolmatch BL, Dong YH, Trerotola SO, et al: Tissue response to covered Wallstents. J Vasc Interv Radiol 9:471-478, 1998.

22. Diamond DA, Vesely TM: The role of radiation therapy in the management of vascular restenosis. Part I. Biologic Basis. J Vasc Interv Radiol 9:199-208, 1998.

Dr. Dolmatch is section head of Vascular/Interventional Radiology at the Cleveland Clinic Foundation in Cleveland, OH.