Arterial Kink and Damage in Normal Segments of the Superficial Femoral and Popliteal Arteries Abutting Nitinol Stents

Abstract

Objective. To identify a new angiographic parameter associated with poor short- and long-term outcomes with nitinol stenting in the larger infrainguinal arteries. Background. Nitinol stents have proven to be useful and safe, but imperfect, tools for treating claudication and limb-threatening ischemia. Primary and secondary patency in superficial femoral artery (SFA) occlusions treated with nitinol stents are up to 80% at 1 year, but restenosis is between 40–50% at 2 years. The causes of SFA and popliteal restenosis remain unclear. Stent fracture has been implicated in some cases of restenosis, but this is clearly the minority. Chronic mechanical trauma to the arteries caused by native vessel-stent interaction, intensified by limb motion over time, appears to be a more plausible explanation. Methods. Presented here are 2 cases of restenosis apparently caused by acute and chronic trauma to the native vessel from interaction of the artery with the ends of relatively rigid nitinol stent systems. Results. The source of some future restenotic and occlusive events are not apparent using routine angiography techniques. Conclusion. The additional step of an on-table leg bend test at 80–90 degrees will allow the interventionalist to visualize many cases of negative interaction between the native artery and the stents that will occur during routine movement. This allows the operator to potentially avoid stent-induced arterial trauma.

Vascular Disease Management 2005;2(5):160–164

Key words: peripheral, artery, stent, restenosis, angiography

The superficial femoral artery (SFA) and popliteal arterial systems are among the last of the large arterial trees to be plagued by clinically excessive restenosis rates. The SFA and popliteal are difficult arteries to treat endovascularly. They are generally diffusely diseased, often occluded, and lie in an area of the body subject to substantial motion. The SFA and popliteal undergo multiple stresses including extension, contraction, compression, elongation, flexion and torsion. Restenosis rates with percutaneous transluminal angioplasty (PTA), surgical endarterectomy and early atherectomy devices, including laser, were unacceptably high.^4,15,16,22

Initial experience with stainless steel stents and Wallstents (Boston Scientific, Natick, Massachusetts) were even worse.^25,27,28 Nitinol stenting in the larger infrainguinal arteries has proven to be a useful, safe, but imperfect, technique for treating claudication and limb-threatening ischemia.^9,10,21–23 Primary and secondary patency in SFA occlusions treated with nitinol stents are up to 80% at 1 year, but restenosis is between 40–50% at 2 years.^18–20

Initial SFA drug-eluting stent (DES) studies were successful in inhibiting restenosis at 6 months, but not in significantly lowering the long-term rate.⁷ Evaluation of nitinol stents in vitro and in animal models showed minimal inflammation and excellent tolerance of nitinol by arterial tissues.^1–3 Nickel elution levels in tissues are not significantly different with nitinol versus stainless steel stents.^1–3

Since the problem of restenosis has been substantially decreased with DES on stainless steel platforms, the causation of higher restenosis rates in SFA and popliteal arteries remains uncertain. Are the causes of SFA/popliteal restenosis the same as those in other arterial trees? Stent fracture has been implicated in some cases of restenosis, but this is clearly the minority. A more ubiquitous process is likely. Chronic mechanical trauma caused by native vessel-stent interaction due to present generation stent design seems more plausible. Damage to the stented SFA/popliteal accumulates with limb motion over time.

Presented here are two cases of chronic restenosis apparently caused by acute and chronic trauma to the native vessel by interaction with the ends of relatively rigid, predominantly nitinol stent systems. This interaction was not apparent using routine angiography techniques; these cases were diagnosed with the additional step of an on-table leg bend test. Bending the procedural leg 80–90 degrees will allow the operator to visualize many cases of negative interaction between the native artery and the stents placed. Bending the artery, pre-stent placement may allow the operator to avoid stenting in areas of the artery prone to excessive movement. These two cases include one acute limb occlusion caused by traumatic native vessel-stent edge interaction. It was contributed to by altering the natural bend in the artery during the stenting process and one chronic restenotic process with similar causation. The length of stent needed to achieve acceptable results with minimal gradient in the SFA/popliteal system is often 17–20 cm.⁸ The proximal and mid-segments of the SFA/popliteal are relatively immobile, but the portion of the artery after it exits the adductor hiatus into the popliteal is not well supported by external tissues and undergoes substantial bending and twisting.^5,6 The areas of bend are not easily identified during straight leg angiography. It was generally assumed that arterial bend occurs in the area of the joint space. We have found this to rarely be the case. The portion of the artery that exhibits the most motion is the area just distal to the adductor hiatus. Stenting across this area of bend with relatively stiff nitinol stents (stents that do not compress well longitudinally) can exacerbate the bend in the native artery at the ends of the stent and cause traumatic interaction between the native artery and the stent edge.

Case 1. A 69-year-old diabetic female presented at our institution with a history of coronary artery bypass graft surgery (CABG), multiple percutaneous transluminal coronary angioplasty (PTCA) procedures and percutaneous transluminal angioplasty (PTA) procedures of bilaterally totally occluded popliteal arteries. The initial PTA of the left popliteal was in May 2002 at which time no stent was placed. The occlusion was approximately 5 cm long over the joint space in the left popliteal artery. The rest of the SFA and popliteal were diffusely, but not critically, diseased. The patient presented again with clinical restenosis and total occlusion of the popliteal in December 2002, at which time area ablative excimer laser atherectomy and angioplasty were performed, and a 6 mm x 30 mm Protégé stent (ev3, Inc., Plymouth, Minnesota) and 7 mm x 80 mm Bridge SE (Medtronic, Inc., Santa Rosa, California) were placed across the distal popliteal and proximal popliteal/SFA, respectively.

She presented again in October 2003 with clinical restenosis on the left. Again, area ablative laser treatment and angioplasty were performed. A 7 mm x 80 mm Protégé stent was placed over the previous 7 mm x 80 mm Bridge SE stent to treat a suboptimal angioplasty result, and a 7 mm x 20 mm Protégé was placed proximally in the mid-SFA due to new disease. During this angioplasty of the left leg, we performed an angiogram of the right popliteal artery where a 5 mm x 40 mm coil stent (ev3, Inc.) had been placed in April 2002 for a short total occlusion of the right popliteal artery. This right popliteal total occlusion was very similar to the original disease seen in the left popliteal artery. The right popliteal coil stent did restenose. It was lasered and re-angioplastied in October 2002.

At the time of the angiogram in October 2003, it showed less than 50% disease. The right popliteal artery has not clinically restenosed as of January 7, 2005. In May of 2004, 7 months after the previous angioplasty and stenting of her left SFA and popliteal artery, the patient presented with acute occlusion of the left leg and limb-threatening ischemia. We performed mechanical thrombectomy using the AngioJet® system (Possis Medical, Inc., Minneapolis, Minnesota), atherectomy using the Silverhawk system (Foxhollow), angioplasty, and stenting of the distal SFA and popliteal artery with a 7 mm x 150 mm Protégé stent. This stent was placed inside the previous 6 mm x 30 mm and 7 mm x 80 mm stents. The previously placed stents were examined with single-frame digital subtraction angiography to rule out stent fracture, and none was found. The initial result appeared excellent, with a normal 3+ dorsalis pedis pulse on discharge.

The patient was discharged on aspirin 325 mg and clopidogrel 75 mg after having been loaded with 300 mg in the hospital. The patient was compliant with her medications. Two days later, the patient presented with a similar acute occlusion of the same leg. Mechanical thrombectomy was performed, and once again, the vessel seemed widely patent on angiography. Intravascular ultrasound (IVUS) was performed on the SFA/popliteal and tibial-peroneal trunk into the anterior tibial. The anterior tibial was her only patent infrapopliteal vessel. There was clear dissection and narrowing of the SFA just proximal to the first 7 mm x 150 mm stent in the distal SFA. There was also dissection distal in the tibial-peroneal trunk. The short 7 mm x 20 mm stent previously placed in the mid-SFA and proximal to the occluded area had mild restenosis only. It was felt at the time that these dissected areas must have been the nidus for thrombus formation, and a short 7 mm x 20 mm Protégé stent was placed over the dissected area proximally, and a short 3.0 mm x 18 mm Cypher™ stent (Cordis Corporation, Miami, Florida) was placed distally at the junction of the tibial-peroneal trunk and the anterior tibial. The proximal end of the Cypher stent was flared to 4 mm. Repeat IVUS showed a good result.

The patient was loaded on coumadin to therapeutic levels and was prescribed aspirin 325 mg. Four days post-discharge, the patient again presented with a similar limb-threatening occlusion of the same leg. The patient’s International Normalized Ratio (INR) upon arrival at the hospital was low, at 1.6. The patient told an interesting story that she was using a wheelchair around the local Walmart for about an hour. At that point she stood, started walking, and shortly thereafter began experiencing limb pain. Again, mechanical thrombectomy left us with a widely patent artery and a mystery as to why the patient continued to occlude. The low INR likely contributed to this, but we did not think that it could explain the entire situation. We did a leg-bend test on the table to 80–90 degrees and observed obvious kinking of the native artery at the proximal and distal ends of the stented region. The stents remained rigid and did not compress longitudinally. The native artery at both ends was clearly being damaged as it was forced against the rigid end of the stents. We did not feel at the time that additional stenting would improve the situation. We decided to recommend against the patient bending her leg, hoping to allow healing of the artery and time to arrive at a more permanent solution. The leg remained patent for 4 months before re-occluding, at which time thrombectomy and atherectomy were again aggressively used to re-open the artery to the distal anterior tibial.

Interestingly, at the time of the most recent procedure, the patient for the first time no longer experienced limb-threatening ischemia from the occlusion. The ankle brachial index (ABI) on the left during occlusion had improved to 46 through the growth of collaterals. This event left us much more concerned about the issue of arterial bending and the effect on that bending exerted by longitudinally rigid stent systems. The difference in clinical course between the two very similar limbs, one treated with longitudinally rigid nitinol (and eventually a rigid Cypher stent distally), and one with a malleable coil stent, was striking. We started routinely performing leg-bend tests on SFA/popliteal angioplasty cases, and have done over 40 at the time of this writing. What we have found is that the area of bend in the proximal popliteal is variable, presumably varying with the location of the adductor hiatus, which has been previously described.

Case 2. A 77-year-old female first presented in November 2001 with rest pain and 5 cm of subtotal disease in the right distal SFA and popliteal; the rest of the SFA had moderate 50% disease. We initially tried PTA only, which appeared successful and resulted in a normal dorsalis pedis pulse and relieved symptoms. The patient returned in August 2002 with severe restenosis of the previous PTA site and new disease proximally. Laser atherectomy was performed on the restenotic and new disease areas. PTA with stenting, using a 6 mm x 100 mm Dynalink (Guidant Corporation, Santa Clara, California) achieved a very good result. The patient did well until January 2003, when she presented with worsening disease in the majority of the SFA proximal to the previous stent. The stented area had less than 50% restenosis. Laser atherectomy and PTA were once again performed on the area of new disease. Stenting with a 7 mm x 100 mm Dynalink and 7 mm x 30 mm Protégé stents left a widely patent lumen.

This result lasted 1 year until January 2004, when she was found to have a total occlusion of the distal end of the distal stent (Dynalink) and a 90% restenosis of the distal end of the proximal stent (Protégé). We again looked at the stents for signs of fracture, but none were found. A suboptimal angioplasty result led us to place a 7 mm x 150 mm Protégé stent in the old 6 mm x 100 mm Dynalink stent, extending distally into the mid-popliteal artery. A normal pulse at the dorsalis pedis was again found. This lasted 7 months, and in August 2004, she was found to have a total occlusion of the distal stent (now 7 mm x 150 mm Protégé). The proximal stents had less than 50% restenosis. Laser (2.5 mm) and Silverhawk atherectomy were performed. At this point, the wire was withdrawn, and an 80–90 degree leg-bend test was performed, showing dramatic flow-limiting kinking of the native popliteal against the distal stent. This was not evident on the straight leg angiogram. A 7 mm x 40 mm coil stent was placed at the distal end of the stent into the popliteal artery and a leg-bend test was repeated, showing a smoother, less obstructive bend in the popliteal artery. This angioplasty result has remained clinically asymptomatic as of January 2005.

Conclusion

Native artery trauma caused by the motion of the SFA and popliteal arteries against the ends of nitinol stents appears to contribute to some cases of acute occlusion, as well as to late restenosis. Some stent designs minimize this effect by allowing longitudinal compression and more gentle bending of the artery with motion. Longer and multilayered stenting increase the likelihood of traumatic arterial stent interaction. Performing a leg-bend test to 80–90 degrees before stent placement may allow the operator to avoid stenting areas prone to excessive motion. Performing a leg bend test to 80–90 degrees after stent placement may allow the operator to identify areas of native artery likely to be traumatized by stent edge interaction during normal patient movements. In areas of excessive arterial bend, the operator may diminish the likelihood of native vessel damage by avoiding the placement of new stents inside previously placed stents. This damage is due to kinking of the native artery against the edge of a longitudinally rigid stent system during normal leg motion. Stenting the proximal and mid-SFA generally did not severely alter the natural bend distally in the popliteal artery, but on occasion, it does intensify the bend. In one case, the loss of “slack” in the proximal vessels caused enough redundancy in the popliteal to cause flow limitation during leg bend and necessitated treatment. We have also noticed a tendency for a gentle “S” bend in the proximal SFA to occur with stenting of the mid-SFA. We have never observed a flow-limiting bend in these cases.