Case Description. A 45-year-old male with established coronary artery disease was admitted for 2 weeks of recurrent rest-onset chest pain symptoms. The patient had undergone 2 previous percutaneous coronary interventions at another hospital. In 1997, an acute stent thrombosis of a second obtuse marginal necessitated urgent coronary artery bypass graft surgery (CABG) with vein grafts to the marginal and right posterior lateral artery. Four months prior to the current admission, he underwent a repeat CABG with a left internal mammary artery (LIMA) graft placed to the left anterior descending coronary artery (LAD). His cardiovascular risk factors included active smoking, hyperlipidemia, hypertension and a family history of premature coronary artery disease. He was taking aspirin, metoprolol and felodipine. Patient management. Following admission, serial cardiac enzymes were negative for myocardial necrosis. A transthoracic echocardiogram demonstrated mild inferior hypokinesis with an overall preserved left ventricular function. Subsequently, the patient underwent a diagnostic cardiac catheterization revealing significant known three-vessel native coronary artery disease. The vein grafts were patent, but there was a severe 95% stenosis of the LIMA at its anastomosis (Figure 1A) with Thrombolysis in Myocardial Infarction (TIMI) grade 2 flow. During the catheterization, the patient experienced severe chest pain with anterior T-wave changes. An immediate percutaneous intervention was performed. Using a 6 French VB-1 guiding catheter (Cordis Corporation, Miami, Florida) to engage the LIMA, the lesion was dilated with a 2.0 x 15 mm Maverick balloon (Boston Scientific/Scimed, Inc., Maple Grove, Minnesota) over a 0.014´´ Choice PT wire (Boston Scientific/Scimed, Inc.). A 2.25 x 18 mm BX Velocity stent (Cordis Corporation) was subsequently implanted at 12 atmospheres because of significant recoil post-balloon angioplasty. Final arteriogram (Figure 1B) demonstrated normal flow with a minor residual stenosis. The patient was pain free at the end of the procedure and his electrocardiogram had normalized. He had an uncomplicated in-hospital course and was discharged the next day on aspirin, clopidogrel, simvastatin, ramipril, metoprolol and a nitroglycerin patch. On the same day as discharge, the patient returned to the hospital with recurrent chest pain symptoms. There were no electrocardiographic changes or elevated biochemical markers consistent with myocardial injury. In view of the temporal relationship of his symptoms to his recent percutaneous intervention, selective arteriography of his LIMA was performed. An acute bend was visualized in the body of the recently implanted stent with resultant severe luminal narrowing (Figure 1C). This was not present after review of the previous interventional final arteriogram. Furthermore, this bend was dynamic, straightening with inspiration (Figure 1D), and varying with the cardiac cycle. This finding was therefore believed to be secondary to mechanical bending created by heart movement. Considering the recent redo-CABG and the acuity of this event, the “bent stent” was reinforced with repeat stenting using a 2.5 x 9 mm NIR Elite stent (Boston Scientific/Scimed, Inc.) deployed at 14 atmospheres. The stent could only cross the lesion during patient-held inspiration. The final arteriogram (Figure 1E) demonstrated a near-normal appearing result with disappearance of the acute bend. The procedure was uneventful, and the patient was discharged the next day after serial negative cardiac enzymes and resolution of symptoms. Unfortunately, the patient was readmitted 3 weeks later with the recurrent symptoms. Repeat cardiac catheterization revealed a bend at the mid-portion of the overlapped stents (Figure 1F). The remainder of the coronary graft anatomy was unchanged. After discussion among the surgical and interventional teams, further revascularization was deferred and the patient was treated medically. How Would You Manage This Case? Gregg W. Stone, MD Cardiovascular Research Foundation New York, New York This is a fascinating and, to my knowledge, unique case in which a 2.25 mm Cordis BX Velocity stent was implanted in a lesion in a relatively straight segment of a LIMA, only to symptomatically recur in a "bent" configuration. Though reinforcing the stent from within with a 2.5 mm Boston Scientific NIR acutely resulted in an improved angiographic appearance, symptoms recurred weeks later, along with the “bent” stent. As a patent LIMA represents the patient’s lifeline, there clearly is nothing “kinky” about this case, appearances notwithstanding. The discussion should focus on the origin of the defect, whether symptoms are indeed attributable to this lesion, and what long-term management strategies might be considered. While the longevity of the LIMA in most patients is well appreciated, less often discussed are problems that the interventionalist (and his/her patient) may encounter when this graft becomes diseased, either early after bypass (usually at the distal anastomosis, related to technical factors, as in the present case), or later on, at which time lesions may present at any location. Poor guide support is often an issue, as well as the tendency of the mammary to develop severe and at times refractory spasm. Perhaps relevant to the present case, when the LIMA is left excessively long, or redundant, in relation to its distal anastomosis to the LAD (a surgical choice arising from the desire to avoid excessive graft shortening during mediastinal healing), interventional “challenges” may result, the most common of which is difficulty in passing guidewires or angioplasty equipment through marked tortuosity in the bypass conduit. In the present case, a stent implanted in a relatively straight segment of the LIMA (at least in the view we’re shown) mysteriously “bends”, presumably made possible by the fact that the LIMA was redundant and mobile with the respiratory cycle, and not yet fibrosed to the surrounding chest cavity. But how can a metallic stent bend? Perhaps this would be explicable for an articulated stent, such as the Palmaz-Schatz, but it is more difficult to envision for a stent with a unicellular design such as the BX Velocity. Clearly, although we think of stents as rigid and resistant to recoil, they were never meant to overcome the flexing and torsional forces that are present in the current case. It is also possible that one or more struts may have fractured, creating a “pseudo-Palmaz-Schatz stent”, complete with articulation (though we are not told of balloon rupture during deployment). However, a second stent implanted at the pivot point did not cure the problem, arguing against initial strut fracture as the cause. In retrospect, one could contend that to better resist recoil and bending, a stent with thick struts, such as the Guidant Ultra, might have been used rather than the thin strut NIR for reinforcement. Of note, both after the original and the additional stents were implanted, a mild residual stenosis was still present in the mid stent at what ultimately becomes the hinge point. Perhaps this area represents a region of marked stent underexpansion, possibly due to a fibrotic area within the graft. Intravascular ultrasound would be useful and is indicated to determine whether an anatomic or structural problem is present at the hinge point. Marked stent under-deployment might have also occurred if the stent slipped off the balloon during passage, a condition that would also be revealed by IVUS, although I would consider this possibility less likely in this case. How should this problem be “fixed”? The operators have elected medical treatment at this time. Of concern, however, is that in Figure 1F; not only does the original stent segment appear significantly narrowed, but a second stenosed region is developing, possibly proximal to the stents. Certainly, if there is any question whether the symptoms are indeed related to the bent stent, a functional test for ischemia should be performed (preferably with adenosine rather than exercise, to minimize the risk of stent thrombosis), or assessment of fractional flow reserve. If IVUS demonstrates stent underexpansion, higher pressure, larger balloons or other measures may be indicated. My favored solution would be a long stent (e.g., 28–38 mm), to provide sufficient straightening force at the ends to inhibit the short stented area from kinking so sharply. I would also recommend long term clopidogrel, in addition to aspirin. Finally, it is possible that as fibrosis develops over time, the patient (or at least his artery) will “straighten himself out.” H. Vernon Anderson, MD University of Texas-Houston Houston, Texas This case illustrates the repetitive bending of a stent placed in a LIMA graft to the LAD. The mechanical forces responsible for this repeated bending likely are related to vertical excursion of the heart in the setting of a relatively fixed LIMA (which is less mobile being attached to the thoracic wall). Reinforcing the first bent stent with a second, supporting stent, did not alleviate the problem. The difficult aspect to me is that this relatively young man has an essentially normal left ventricle with a jeopardized anterior wall. In my mind, the likelihood that the combined bent stents will ultimately either suffer acute thrombosis or enhanced neointimal proliferation is quite high. In the situation presented here, I believe that the sudden and catastrophic loss of the LIMA is probable. Although the prospect of a repeated bypass operation is very unattrative, I believe that my approach would be to suggest that bypass surgery be presented to the patient, envisioning placement of a saphenous vein graft to the distal LAD. It might also be possible to revise the LIMA-LAD anastomosis; however, I do not believe that this would be the most feasible solution. There may, of course, be other factors that mitigate against a repeat bypass. For example, severe lung disease or other systemic diseases might be factors. However, if such is not the case, and since percutaneous intervention has not solved the problem, then consideration of another bypass is necessary in order to help preserve this young man’s left ventricular function. Satish Surabhi, MD and Sheldon Goldberg, MD MCP/Hahnemann University Hospital Philadelphia, Pennsylvania This case demonstrates a rare situation which we have not encountered in over 12 years of stent implantation — namely, mechanical bending of a stent due to cardiac motion. If excessive torsion of the stent could be identified before the initial stent placement, one could possibly avoid this situation by placing a flexible nitinol self-expanding stent such as the RADIUS™. This might have adapted better to the local mechanical environment as compared with a rigid scaffold, such as was used on 2 occasions in this case. Since the placement of 2 stents has failed, medical management is certainly a reasonable alternative option. However, if symptoms recur, we believe another possibility would be to perform an interposition graft with the use of the radial artery as a conduit around the stent. This might even be able to be done off-pump. We would certainly obtain a straight lateral (90º left anterior oblique) view of the course of the LIMA to be sure it is not running directly underneath the sternum prior to doing any further surgical intervention. Howard Cohen, MD Presbyterian University Hospital Pittsburgh, Pennsylvania This case represents a very unusual and difficult problem to treat. The problem had not be present immediately post-operatively but may develop as adhesions develop and the LIMA graft shortens or as the lesion progresses at the anastomosis as the result of kinking in this area. Furthermnore, the problem may not always be readily apparent angiographically particularly if the angiogram is done during full inspiration which tends to elongate the graft and relieve the kinking. Treatment alternatives include insertion of a stent with increased radial strength as was done in this report with insertion of a NIR stent. Another alternative would be the insertion of a “self-expanding” stent such as a Wallstent that should be more resistant to kinking or crush. For this reason, this type of stent has been favored in carotid stenting to avoid crushing of the stent by external pressure. Finally, the possiblity of PCI in the native vessel should always be reconsidered, but this may not always be possible. In some cases, the situation may not be resolvable with percutaneous intervention or medical therapy in which case re-do CABG may be required if the amount of myocardium in jeopardy is great.