Reduction of Subacute Stent Thrombosis (SAT) Using
Heparin-Coated Stents in a Large-Scale, “Real World” Registry

ABSTRACT: Purpose. This study was designed to compare the rates of subacute stent thrombosis (SAT) among patients receiving heparin-coated stents to patients receiving bare-metal stents in real world, contemporary coronary interventions. Background. Controlled trials with heparin-coated coronary stents have shown a trend toward decreased rates of SAT. Methods and Results. The data in this study were collected from a single, large cardiac center over a period of 9 months. All patients who underwent coronary stent implantation during this 9-month period were included in the study (1,288 patients; 1,366 procedures; 2,231 stents). All patients were treated with aspirin and clopidogrel (or ticlopidine) after stenting. Primary thrombotic outcome was defined as angiographically documented SAT and/or sudden unexplained cardiac death (SCD) within 30 days of the procedure. Follow-up data (1,264/1,276 patients) were obtained in 99% of patients. A total of 337 patients received 543 heparin-coated stents (BX Velocity™ Hepacoat™) and 939 patients received bare-metal stents (1,688 stents). SAT was seen in 25/1,024 procedures (2.44%) in the bare-metal stent group and 1/342 procedures (0.29%) in the heparin-coated stent group. Primary thrombotic outcomes (SAT or SCD) were observed in 31/1,024 procedures (3.03%) in the bare-metal stent cohort and in 2/342 procedures (0.58%) in the heparin-coated stent group. The vast majority (96%) of the patients who had SAT within 30 days had initial stent placement for an acute coronary syndrome (p 1,2 Most of these events take place in the first month of stent placement.³ A number of factors contribute to the formation of thrombus after stent deployment. Stents are made from metallic compounds that have inherent thrombogenicity and can often lead to stent thrombosis.^4,5 Optimal stent placement is an important factor in preventing SAT, as shown by Colombo et al.⁶ There appears to be an increased risk of SAT in patients with pre-existing thrombus, dissection or multivessel disease, and acute coronary syndromes.^5,7–10 Even with adequate technique and dual antiplatelet therapy, stent thrombosis is observed in 1–3% of patients after coronary stenting with bare-metal stents.^12–23 The clinical manifestations of SAT are severe, and include death in 25–50% of patients and major myocardial infarction (MI) in 40–70%.^{12,14,23–26} Local antithrombin therapy with heparin-coated stents is a potentially attractive means to limit this morbid complication. The enzymatic nature of the Cardmeda™ heparin coating allows for persistent thrombin inactivation at the stent surface. A number of studies have shown a trend toward decreased SAT rates with the use of heparin-coated stents.^27–32 Despite the consistent reduction of SAT rates with heparin-coated stents in these studies (compared to historical controls), to date, there have been relatively few data presented to assess the efficacy of heparin-coated stents to reduce the incidence of SAT in a “real world,” every-day stenting. In this large, single-center registry, we attempted to address this question by comparing the incidence of documented SAT in bare-metal versus heparin-coated stents in 1,688 consecutive, real world stent procedures (2,244 stents). Methods The study was a non-randomized, retrospective, large, single-center registry of all patients who underwent a stent implantation of 1 or more coronary arteries, from November 2000 through July 2001, at the Medical Center. The primary endpoint(s) for the study were SAT and/or unexplained sudden cardiac death within 30 days of their stent procedure. Thirty-day follow-up data (1,264/1,276 patients; 99%) were obtained by telephone contact, chart review, angiographic case review and review of hospital records. The patient population treated with bare-metal stents was compared to those patients receiving 1 or more heparin-coated stent. Inclusion and exclusion criteria. Inclusion criteria were kept very broad so as to represent true real world stenting. Every patient who was hospitalized and underwent placement of a coronary artery stent, either electively or as an emergent procedure for management of unstable angina or acute MI, was included. Stent type was chosen at the discretion of the treating physician. Patients with clinical and/or angiographic evidence of thrombus in the coronary artery as well as those patients with coronary dissection who underwent stenting were all included. Patients were excluded from the study if a stent was not deployed due to technical difficulties, if angioplasty was done without any stent placement, if the patient died during the stent deployment procedure, or if the patient had both heparin-coated and bare-metal stent(s) placed during the same procedure. Patients treated with warfarin and/or aspirin only (no clopidogrel or ticlopidine) after stenting were excluded (n = 10). Data collection and follow-up. An experienced physician, using standardized data collection forms, abstracted medical records of all patients who underwent stent deployment procedures. During the chart audit, procedures were divided into bare-metal stent or heparin-coated stent groups. Data were collected with regard to other potentially confounding variables, including stent diameter, stent length, number of stents implanted, final balloon size, maximum inflation pressure, indication(s) for intervention, bare-metal stent type, operator implanting stent(s), anticoagulation regimen after stenting and glycoprotein (GP) IIb/IIIa inhibitor use. A research assistant and a registered nurse contacted all patients or their family members via telephone and asked standardized questions about hospitalization and events relating to their cardiac status within 30 days of the procedure. All occurrences of coronary events and deaths in that standard (SAT) time frame were recorded. The hospital’s electronic data system was also used to track emergency department visits and admissions for all patients within 30 days of the initial procedure and charts were audited for possible major coronary events. If the patient had a repeat angiogram or stent placement within 30 days, the angiogram was reviewed for possible SAT. Two blinded, experienced film readers served to adjudicate the classification of the follow-up angiogram as an SAT (see definitions below). There was no intent in the study design to assess late events (> 30 days after stent implantation). Patients who underwent additional stent placement procedures within 30 days of the initial procedure were followed for an additional 30 days. Study population. A total of 1,332 patients were included in the study cohort. Sixty-four patient records were excluded from the study. Of these, twenty-nine had both heparin-coated stent(s) and bare-metal stent(s) placed during the same procedure and 13 had both stent types placed within 30 days of the initial procedure. Twelve patients were lost to follow-up (11 in the bare-stent group, 1 in the heparin-coated stent group). Ten patients were prescribed either warfarin or aspirin only after stenting, and were excluded (no events in this cohort). The remaining 1,276 patients had 1,366 stent placement procedures. Patients were grouped according to the stent type used during the index procedure, i.e., the bare-metal stent group (n = 939 patients with 1,024 procedures) or heparin-coated stent group (n = 337 patients with 342 procedures). Antiplatelet and anticoagulation therapy. Patients received 325 mg aspirin before the intervention and every day thereafter. Clopidogrel was administered as a 300 mg loading dose either before or immediately following the procedure, followed by 75 mg per day for 4 or more weeks.^33–36 If clopidogrel was contraindicated, then ticlopidine was administered as a 500 mg loading dose either before or immediately following the procedure and then at 250 mg twice daily for a minimum of 4 weeks.^33,35,36 Nine bare stent patients had clopidogrel without aspirin (no thrombotic events), and 3 patients took ticlopidine due to side effects or allergies to clopidogrel (no thrombotic events). Heparin was administered during the procedure either as a standard bolus of 10,000 I/U or according to a weight-adjusted protocol (70–100 I/U per kg). Additional heparin was given during the procedure depending on the activated clotting time (ACT). ACT was generally measured every 30–60 minutes and maintained at > 200 seconds if GP IIb/IIIa inhibitors were used and at > 250 seconds otherwise.³⁷ Heparin was typically discontinued after the procedure. Intravenous GP IIb/IIIa inhibitors were administered at the discretion of the operating cardiologist. The GP IIb/IIIa inhibitors administered were either abciximab or eptifibatide (Table 1). The dosage was prepared according to the weight-adjusted protocol. Percutaneous coronary intervention. Stent implantation and balloon angioplasty were performed according to the standard clinical practice by the femoral approach. Implantation was usually done after balloon predilatation. Direct stenting was performed in 32% of cases. Multivessel stenting was performed in 47% of the patients. A very small number of patients had cutting balloon angioplasty (n = 14) or rotational atherectomy (n = 21). The bare-metal stents used in this study included the Bx Velocity™ stent (n = 812; Cordis Corporation, Miami, Florida), AVE™ stents (n = 442; GFX2, S670, S660, Medtronic AVE, Inc., Santa Rosa, California), the Multi-Link Tetra™ stent (n = 174; Guidant Corporation, Temecula, California) and the NIR™ stent (n = 260; Boston Scientific/Scimed, Inc., Maple Grove, Minnesota). The Bx Velocity™ Hepacoat™ stent was used in all of the heparin-coated stent cases (n = 543). Stent type, size and length were chosen at the operating cardiologist’s discretion. Routine high-pressure (> 12 atmospheres) post-dilatation was done in every case. Thirteen different interventional cardiologists from 4 different practice groups implanted stents. Endpoints and definitions. The primary endpoint of the study was angiographically documented SAT within 30 days of stent implantation. The secondary clinical endpoint was defined as a composite of angiographically documented thrombosis of the stent(s), MI without follow-up angiography and/or unexplained cardiac death within 30 days of the initial stent deployment procedure. MI was not regarded as a thrombotic endpoint if the stent placed during the earlier procedure was found to be angiographically patent at follow-up. SAT was defined as either angiographically documented complete occlusion (TIMI flow 0 and 1) or a flow-limiting intracoronary thrombus (TIMI flow 1 or 2)^27,28 associated with a clinical presentation of angina and/or MI. Intracoronary thrombus was defined as the presence of a filling defect within the lumen surrounded by contrast material or the persistence of contrast material as seen in multiple projections in the absence of calcium within the filling defects. The important endpoint of SAT was adjudicated by consensus between 2 blinded, experienced film readers. Death was categorized as cardiac, non-cardiac or unexplained. Cardiac death was defined as any death related to cardiac events, such as cardiogenic shock, myocardial rupture or coronary artery dissection. Death within the hospital with a specific diagnosis that was clearly unrelated to cardiac disease was regarded as non-cardiac. Any unexplained death outside the hospital or sudden death within the hospital without known etiology was referred to as unexplained (presumed cardiac) death. MI was diagnosed if there were new pathologic Q-waves according to the Minnesota code,38 or if there was an increase in the serum creatinine phosphokinase to greater than twice the upper limit of normal value together with a pathological increase in concentration of myocardial isoenzymes (MB > 8%). Statistical analysis. All continuous variables are described as means ± standard deviation and compared between groups by student’s t-test for means. Gender between the 2 groups was analyzed using the student’s t-test for proportions. Categorical variables were described as counts and percentages and compared between the groups using 2 x 2 contingency tables for Mantel-Hansel Chi-square analysis. Additional confirmatory multivariate regression was used to control for the potentially confounding variables of stent number, stent length, nominal stent diameter and acute coronary syndrome. All tests were 2-tailed, with a 5% level of significance (alpha = 0.05). SAS software was used for all analyses.³⁹ Results Clinical characteristics. The study group was comprised of 1,288 patients undergoing 1,366 stent placement procedures. Seventy-eight patients had more than 1 procedure during the 9-month study period. Thirty-day follow-up was completed in 99% of cases (1,276/1,288 patients). The baseline characteristics of the bare-metal and heparin-coated stent groups are summarized in Table 1. There were no significant differences in the clinical characteristics, except for diabetes and atrial fibrillation. Both of these clinical variables were seen more frequently in the heparin-coated stent group (diabetes p-value = 0.03; atrial fibrillation p-value = 0.026). There were no statistically significant differences in average stent size, total stent length, number of stents implanted, maximum balloon size, inflation pressure or GP IIb/IIIa inhibitor use (slightly greater in the bare-stent group) between the bare-metal and heparin-coated stent groups. The indications for stent placement and incidence of stenting for acute MI and acute coronary syndromes were not statistically different between the bare-metal and heparin-coated groups. Multivariate statistical analyses demonstrated that the differences in SAT between the heparin-coated and bare-stent groups were independent of all of these potentially confounding covariables. Clinical outcome. Of 1,366 procedures performed, thirty-three procedures (33 patients) met the primary thrombotic outcome. We found 26 confirmed cases of SAT and 7 cases of unexplained (presumed cardiac) death within 30 days of stent placement. SAT rate for the cohort as a whole was 1.9%. An additional 0.5% of patients suffered an unexplained death within 30 days of stenting. The clinical outcome of patients with SAT was grave. Eighteen of 26 patients (70%) presented with acute coronary syndrome and 3 patients (12%) eventually died due to clinical deterioration. This death rate approaches 30% if the 7 unexplained dead patients are assumed to be caused by SAT. Of the 33 patients who met the primary thrombotic outcome, all except 1 patient (96%; p 19,21 With improved stent deployment techniques utilizing high-pressure balloon dilatation,^6,25 the rate of stent thrombosis dropped to less than 3.5%.^13–16,21 Dual antiplatelet therapy with aspirin and clopidogrel or ticlopidine has further reduced the incidence of SAT.^13,16,23 Presently, the rate of SAT ranges from 1–3% in controlled trials in elective stenting.^{12,13,15,18,20,22,24} Despite these advances in stenting techniques, the elimination of stent thrombosis continues to pose a major challenge. Although it is relatively rare, SAT is a potentially devastating complication that is associated with substantial morbidity, mortality and economic costs.^24,34 Historically, most of the studies reporting the incidence of SAT have been controlled trials in electively treated, single-vessel disease patients. In nearly all of these trials, the procedures were performed by experienced, high-volume operators. As such, these trials may not accurately reflect the true incidence of SAT in the “real world.” Importantly, these studies often excluded patients at a higher risk for SAT, such as those with unstable angina, with an acute or recent acute MI, those with angiographically evident thrombus, stenting of residual dissection, stenting in arteries with diffuse disease, and multivessel stenting. In order to provide a more meaningful examination of the incidence of stent thrombosis, and the potential effect of heparin-coated stents, our study included a large (consecutive) cohort of coronary stent procedures without pre-selection of low-risk patients. In addition, both high- and lower volume operators implanted stents. Heparin, a potent antithrombotic agent, has been used by several manufacturers to coat bare-metal stents. With the advent of newer heparin-coated stents, the incidence of SAT has been reproducibly reduced to less than 1%.^28–32 In the pioneering BENESTENT II study, the rate of SAT was only 0.2% using a heparin-coated Palmaz-Schatz stent.²⁸ The HEPANET registry recently compared heparin-coated stents to bare-metal stents in 1,603 patients. This registry also demonstrated a trend toward decreased SAT rates and target lesion revascularization among patients treated with heparin-coated stents.³² The Bx Velocity™ Hepacoat™ coronary stent surface is coated with the proprietary Carmeda® Bioactive Surface (CBAS) coating agent, coupled with covalently bonded stable heparin molecules. This has been shown to be biostable and provide relatively long-term antithrombotic activity.⁴ Heparin has well established, enzymatic, antithrombotic properties. Heparin inhibits the activation of coagulation via activation of antithrombin III, which inactivates activated thrombin, thereby inhibiting fibrin formation and thrombin-mediated platelet activation at the surface of the stent.^4,27,40 The current study demonstrates a statistically significant reduction of SAT using a heparin-coated stent compared to bare stents in real world coronary interventions. Overall, the incidence of SAT (1.9%) and the primary thrombotic outcome (2.4%) was slightly higher in our study compared to some other controlled studies looking at similar endpoints in simple elective stenting. This discrepancy is best explained by the broad inclusion criteria of the current study. In our registry, fifty-nine percent of the study population presented with acute coronary syndromes, and nearly half of the patients had multivessel stenting. Importantly, an even higher incidence of SAT was observed in the Arterial Revascularization Therapy Study (ARTS) trial, comparing bypass surgery to elective multivessel stenting.^17,18 In the ARTS study, the incidence of SAT using bare-metal stents was 2.8% (per procedure) at 30 days,¹⁷ as compared to the 2.4% incidence in the bare-metal group in our study. Although the ARTS study included patients with multivessel stenting, it excluded patients with acute coronary syndromes and acute MI. In comparison to the ARTS study, the incidence of SAT in the bare-stent arm of the current study appears to be, if anything, lower than one would predict for such a high-risk, all-inclusive cohort. In the COAST trial, which used the Jomed heparin-coated stent, the complication rates were similar in both bare-metal and heparin-coated groups. Thrombotic events at 30 days were seen in approximately 0.5% of both the bare- and heparin-coated groups. It is important to note that this was a small trial, which was intended to look at the effects of heparin coating upon restenosis. This study was not nearly large enough to adequately compare the SAT rates between the 2 cohorts. In addition, patients with acute coronary syndromes were excluded in this study.⁴² Thus, this study should not be misinterpreted as showing a failure of the Jomed, heparin-coated stent to reduce SAT. The recently published HOPE and PHAROAH trials further demonstrate the efficacy of the Carmeda® heparin coating to prevent SAT. In these 2 trials, the Bx Velocity™ Hepacoat™ was implanted using aspirin as the only antiplatelet drug.^43,44 In the STARS trials with elective bare-metal (single-lesion) stenting, the aspirin-only group had a 3% SAT rate. In contrast, in a similar patient cohort, the combined HOPE and PHAROAH results demonstrated only 2 SATs in 347 patients (0.5%) treated with aspirin only. In the current study, there was only 1 documented SAT among the 342 procedures utilizing 1 or more heparin-coated stent (543 stents). This SAT event occurred following suboptimal placement (under-deployment on a bend, jailing of side-branch and possible edge dissection) of a 2.25-mm diameter, 23-mm long Bx Velocity™ Hepacoat™ stent in a tortuous portion of a small, non-dominant right coronary artery (Figure 2). This patient came back with chest pain after 7 days and was found to have a subtotally occluded stent. Stent thrombosis was much more frequent in the bare-stent group. In contrast to the case shown in Figure 2, the bare-stent cohort included numerous SAT events following (angiographically) optimal placement of larger-diameter stents in vessels with good runoff (Figure 3). In the setting of good stent implantation technique and appropriate antithrombotic and antiplatelet therapy, acute coronary syndrome has become one of the most potent predictors of SAT. One of the important observations from our study was that 96% percent of the cases with a primary thrombotic outcome presented initially with an acute coronary syndrome. In patients presenting with an acute coronary syndrome, there was a 12-fold reduction in SAT risk in patients treated with the heparin-coated stent compared to a bare-metal stent (p Study limitations. This was a retrospective, non-randomized trial. However, the large number of patients in the 2 study groups and the lack of any clear difference in the pre-procedural risk of SAT (bias) between the 2 groups suggest that the results observed are clinically relevant. Since we only obtained follow-up data out to 30 days after stenting, stent thrombosis events that occurred beyond 30 days were not captured in our analysis. Finally, we may have overestimated the incidence of stent thrombosis by including patients with unexplained death as a presumed stent thrombosis. However, this was done in both groups and most prior studies analyzing stent thrombosis have applied similar study designs.^12,14Conclusion. This large, single-center registry demonstrates a substantial and statistically significant reduction of SAT using heparin-coated stents when compared to bare-metal stents in an unselected population of patients treated with coronary stenting. This benefit of heparin-coated stenting was most apparent in patients treated in the setting of acute coronary syndromes.