Author + information
- Received February 18, 2002
- Revision received June 14, 2002
- Accepted July 3, 2002
- Published online November 6, 2002.
- James L Orford, MBChB, MPHa,
- Ryan Lennon, MSa,
- Steven Melby, RNa,
- Panayotis Fasseas, MDa,
- Malcolm R Bell, MBBS, FRACPa,
- Charanjit S Rihal, MD, FACCa,
- David R Holmes, MD, FACCa and
- Peter B Berger, MD, FACCa,* ()
- ↵*Reprint requests and correspondence:
Dr. Peter Berger, Division of Cardiovascular Diseases, Mayo Clinic, W16, 200 First Street SW, Rochester, Minnesota 55905, USA.
Objectives The study examined the frequency, correlates, and outcome of patients with stent thrombosis within 30 days of stent placement.
Background Patients in trials evaluating stents or dual antiplatelet therapy to prevent coronary stent thrombosis have generally had narrow inclusion criteria; the extent to which stent thrombosis rates in such trials represent current practice, particularly with the availability of newer stents, is unclear.
Methods We performed a retrospective analysis of the Mayo Clinic Percutaneous Coronary Intervention database and identified all patients who received at least one coronary stent and dual antiplatelet therapy (aspirin and ticlopidine or clopidogrel for two to four weeks).
Results Four thousand five hundred nine patients underwent successful coronary stent implantation and were treated with dual antiplatelet therapy between July 1, 1994, and April 30, 2000. Stent thrombosis occurred in 23 patients (0.51%; 95% confidence interval 0.32%, 0.76%) within 30 days of stent placement. Multivariate analysis using bootstrap model selection to avoid over-fitting the model indicated that only the number of stents placed was an independent correlate of stent thrombosis (odds ratio 1.80, p < 0.001). The frequency of death and frequency of nonfatal myocardial infarction (MI) among the 23 patients with stent thrombosis were 48% and 39%, respectively.
Conclusions Stent thrombosis is even more rare in the current era than in earlier trials. Number of stents placed was an independent correlate of stent thrombosis. Most patients who suffer stent thrombosis either die or suffer MI.
Five randomized, controlled trials have established the superiority of dual antiplatelet therapy with aspirin and a thienopyridine when compared with aspirin alone, or aspirin and warfarin, in preventing stent thrombosis after percutaneous coronary intervention (PCI) and stent placement (1–5). However, the frequency of stent thrombosis differed significantly between these trials. Furthermore, these trials applied strict inclusion and exclusion criteria, and the frequency of stent thrombosis in such trials may not accurately reflect the frequency of stent thrombosis in current clinical practice. Presently, the use of ticlopidine has been replaced by clopidogrel, and different stents are used from those utilized in the randomized trials; therefore, the frequency of stent thrombosis may differ from that seen in the prior trials.
We sought to determine the frequency and timing of stent thrombosis and other adverse events occurring within the first 30 days after coronary stent placement.
We performed a retrospective analysis of the Mayo Clinic PCI database and identified all patients who received at least one coronary stent and dual antiplatelet therapy. This computerized database includes prospectively collected baseline, procedural, and angiographic data on all patients undergoing PCI at the Mayo Clinic, Rochester, Minnesota. The indications for PCI included stable angina (elective procedures), unstable angina, and acute myocardial infarction (MI) (urgent and emergency procedures).
Angiographic, procedural, and clinical outcomes were recorded prospectively on line by experienced interventional cardiology or coronary care unit nurses before the patient’s dismissal. All patients were contacted at 6 months, 12 months, and yearly thereafter by a clinical research nurse, according to a protocol approved by the Mayo Foundation Institutional Review Board. Medical records of all patients requiring hospitalization at the Mayo Clinic and elsewhere were reviewed to further characterize any clinical events during the study period. This study was approved by the Mayo Clinic Institutional Review Board.
Inclusion and exclusion criteria
All patients were included in this analysis if they had undergone successful deployment of at least one coronary stent without the occurrence of a major in-lab complication (death, Q-wave MI, emergency cardiac surgery, cerebrovascular accident, transient ischemic attack, coronary perforation, or development of cardiogenic shock). Patients were excluded if there was intent not to administer aspirin or a thienopyridine, if they were in cardiogenic shock before the procedure, if brachytherapy was administered, if coumadin was administered prior to discharge from the hospital after the index procedure, if an experimental fibrin-coated stent was used, if the patient experienced stent thrombosis after 30 days post PCI, or if the patient refused permission for his or her records to be used for research (as required by Minnesota state law). If patients underwent more than one qualifying procedure during the study period, only the first stent procedure was analyzed.
All patients received 325 mg of aspirin before intervention and every day thereafter. A 500-mg loading dose of ticlopidine was administered orally immediately before the procedure, after which ticlopidine, 250 mg twice daily, was administered for four weeks. From 1996 onward, ticlopidine was administered for only two weeks (6). Since September 1998, clopidogrel administered as a 300 mg loading dose immediately before the procedure, followed by 75 mg/day for two or four weeks, was used instead of ticlopidine (7). Heparin was administered on a weight-adjusted basis immediately before the procedure, with an aim for an activated clotting time (ACT) of approximately 300 s (initial dose, 100 U/kg) when a glycoprotein IIb/IIIa inhibitor was not used and approximately 250 s when a glycoprotein IIb/IIIa was given (initial dose, 70 U/kg). The ACT was generally remeasured every 30 to 60 min during the procedure, and additional boluses of heparin were given to maintain the ACT in the target range. Heparin was almost invariably discontinued at the end of the procedure, although in very rare instances at the operator’s discretion it was continued for up to 24 h in selected high-risk patients. The decision to use a glycoprotein IIb/IIIa inhibitor was made by the cardiologist performing the procedure, who considered clinical and angiographic features of the patient and avoided the use of a glycoprotein IIb/IIIa inhibitor if contraindications were apparent.
Coronary angioplasty and intracoronary stent implantation were performed after standard PCI techniques and as previously described (8). Delivery balloons were inflated to nominal pressure. Throughout the study period (July 1, 1994, to April 30, 2000), stent delivery has been routinely followed by high-pressure (≥14 atm) balloon inflations using noncompliant or minimally compliant balloons at sizes equivalent to or slightly larger than nominal stent size to achieve residual diameter stenosis <20% without intraluminal filling defects. Intravascular ultrasound was performed at the discretion of the operator.
Successful stent placement was defined as a residual stenosis within the stented segment of <50% by visual estimation and the ultimate achievement of Thrombolysis In Myocardial Infarction (TIMI) flow grade 3. Stent thrombosis was considered to have occurred when confirmed angiographically (intraluminal filling defect within the stent resulting in TIMI grade 0 or 1 anterograde blood flow), when death was sudden and unexplained, or when an MI (both ST-segment elevation and non–ST-segment elevation MI) occurred in the territory of the treated vessel and stent thrombosis could not be excluded definitively.
All patients were contacted via telephone or mail using a standardized questionnaire at 6 months, 12 months, and then yearly after their procedure. All adverse events were confirmed by reviewing the medical records of patients followed up at our institution and by contacting the patients’ physicians and obtaining and reviewing the hospital records of patients followed up elsewhere.
Data are presented as percentages, and means as ± 1 SD. Comparisons between groups are made using one-way analysis of variance or the Pearson chi-square test. Confidence intervals for event rates are calculated using exact binomial algorithms. A multiple logistic regression model for early stent thrombosis was developed using bootstrap selection methods (to avoid an overfit model). Twenty-one variables that correlated with early stent thrombosis at the 0.15 significance level comprised the full model. (The cutoff of 0.15 was chosen because variables may be significant in a multiple regression model but not in a simple regression model.) Two hundred bootstrap samples were selected; stepwise selection of the 21 variables was applied to each sample. The final model consisted only of variables selected in at least 140 (70%) of the bootstrap models (9). Analyses were performed using SAS software (SAS Inc., Cary, North Carolina).
Four thousand six hundred five patients underwent PCI with successful coronary stent implantation and were treated with dual antiplatelet therapy between July 1, 1994, and April 30, 2000, constituting approximately 50% of all the PCI procedures at the Mayo Clinic, Rochester, Minnesota, in that time period. The most common reason for exclusion from the database was balloon angioplasty treatment without adjunctive stenting (∼30%). Eighty-six patients (1.9%) refused permission for their records to be used for research. Twenty-three patients developed stent thrombosis within 30 days of stent placement (0.51%, 95% confidence interval [CI] 0.32%, 0.76%). Of the patients included in this analysis, all but one (99.98%) were contacted, and their clinical status reviewed. The analysis includes the 4,508 patients who gave permission for their records to be used for research and who were contacted; patients refusing permission for their data to be analyzed were excluded, as is required by Minnesota law.
Baseline clinical characteristics
The mean age of the patients (13 men, 10 women) who developed stent thrombosis was 69.6 ± 15.2 years (Table 1). These patients had a higher incidence of congestive heart failure (CHF) on presentation and were more likely to have a past medical history of CHF or metastatic cancer (p < 0.05) than patients who did not develop stent thrombosis.
Baseline angiographic and procedural characteristics
Univariate predictors (Table 2) of stent thrombosis included left main PCI, number of stents placed, number of segments treated, minimum stent diameter, calcium in any stenosis, American Heart Association type C lesion characteristics, and angiographic evidence of plaque ulceration (p < 0.05). However, the large majority of patients with these characteristics did not develop stent thrombosis.
Multivariate analysis using bootstrap model selection identified the number of stents placed (selected in 84.5% of bootstrap samples) as an independent correlate of stent thrombosis (odds ratio [OR] 1.80, p < 0.001). No other covariate was selected in more than 70% of the samples. When adjusted for number of stents in separate logistic models, CHF on presentation (OR 3.4, p = 0.03), concentric lesions (OR 3.0, p = 0.03), ulcerated lesions (OR 2.8, p = 0.03), history of CHF (OR 2.9, p = 0.04), and metastatic cancer (OR 7.9, p = 0.05) were significantly correlated with stent thrombosis (Table 3). These five variables could be considered good candidates for also being independently correlated with stent thrombosis. Indeed, they were significantly correlated with stent thrombosis in our patients. However, the bootstrap analysis indicates that these relationships may be specific to this patient population.
The frequency of death and nonfatal MI among the 23 patients with stent thrombosis was 48% (95% CI 27%, 69%) and 39% (95% CI 20%, 61%), respectively. The rate of nonfatal ST-elevation or Q-wave MI was 30% (95% CI 13%, 53%). One patient (4%; 95% CI 0.1%, 22%) was identified on the basis of diagnostic angiography alone.
The incidences of death and nonfatal MI within 30 days of stent placement among the 4,508 study patients were 0.8% (n = 38; 95% CI 0.6%, 1.2%) and 4.9% (n = 241; 95% CI 4.3%, 5.5%), respectively.
Among patients without stent thrombosis in the 30 days after their PCI procedure, the one-year Kaplan-Meier mortality rate was 3.1%, the non-fatal MI rate was 7.1%, and the combined death or non-fatal MI rate was 10.1%. The rate of repeat PCI of the treated vessel (in patients alive at one year and free of MI) was 8.1%, and the rate of coronary artery bypass graft surgery was 4.9%. The estimated rate of the composite end point of any major adverse cardiac events (death, non-fatal MI, or target lumen revascularization) was 19.4%.
This study demonstrates that stent thrombosis in the current era of improving stent design and construction and dual antiplatelet therapy is even lower than previously reported in most randomized trials and registry studies. Also, although a number of correlates of stent thrombosis were identified, most patients with these characteristics do not experience stent thrombosis. Most patients who do experience stent thrombosis, however, either die or suffer an ST-elevation or Q-wave MI.
Rationale for the use of dual antiplatelet therapy
Ticlopidine and clopidogrel are thienopyridine derivatives that inhibit platelet aggregation by irreversibly inhibiting the binding of adenosine diphosphate (ADP) to one of three identified ADP receptors, the P2Y12receptor, which limits binding of fibrinogen to the glycoprotein IIb/IIIa receptor complex (10). Platelet aggregation in response to collagen, thrombin, and shear stress, all putative factors in the pathogenesis of stent thrombosis, are also inhibited.
Incidence of stent thrombosis
The initial randomized clinical trials comparing stent implantation with balloon angioplasty revealed improved angiographic and, in one study, clinical outcomes (11,12). This was achieved at the cost of an increase in bleeding complications as a result of the use of aggressive anticoagulation regimens following stent implantation. However, despite such therapy, stent thrombosis rates were approximately 3.5% and associated with significant morbidity and mortality. The routine use of high-pressure stent deployment and dual antiplatelet therapy, including a thienopyridine derivative, led to large reductions in the incidence of this highly morbid complication (13,14). Subsequent registries and randomized controlled trials have documented even lower rates of stent thrombosis (0.5% to 1.9%) and a much lower risk of bleeding than was seen with the prolonged use of heparin while coumadin was being initiated (2–5,15–17). However, these clinical trials generally had narrow inclusion criteria, and the extent to which these trials reflect current clinical practice is unclear. The true frequency of stent thrombosis may also be influenced by substantial improvements in stent designs, potentially rendering the earlier trial and registry data outdated.
The current study
This study documents a very low rate of stent thrombosis in a large, contemporary (years 1995 to 2000) cohort of patients treated with dual antiplatelet therapy in which improved stent technology and interventional techniques were available and utilized. This low rate of stent thrombosis occurred despite the presence of many clinical and angiographic risk factors previously associated with higher frequencies of stent thrombosis. Even among patients with these adverse characteristics, the frequency of stent thrombosis was very low.
Risk factors for stent thrombosis
A number of risk factors for stent thrombosis have been identified by previous investigators. Persistent dissection, longer stent length, and final lumen diameter within the stent were identified as independent multivariable predictors of stent thrombosis in a meta-analysis of 6,186 patients enrolled in six major clinical trials of coronary stenting (17). Other studies have identified balloon size ≤2.5 mm, bail-out situations, unstable angina or acute MI (15), ejection fraction, use of a combination of different stents, postprocedural dissections, and slow flow (16)as predictive of stent thrombosis.
Multivariate analysis of correlates of stent thrombosis is difficult to perform in even a study as large as this because of the low incidence of the primary end point. A common method of avoiding over-fitting is bootstrap model selection. It involves creating random data sets from the patient population, selecting correlates with stepwise model selection for each bootstrap sample, and counting the number of patients in which a correlate was selected. Typically, any correlate selected in at least 70% of patients is chosen for the final model. The only variable that was selected at least 70% of the time was the number of stents placed. Therefore, the only variable for which we have strong evidence of an association with early (30 days) stent thrombosis is number of stents placed. When adjusted for number of stents, CHF, ulcerated lesions, body mass index (per 5 kg/m2), and metastatic cancer were also significant correlates of stent thrombosis.
Clinical outcome of stent thrombosis
The clinical consequences of stent thrombosis are generally severe. Mortality has generally been reported to exceed 20%, and the combined end point of death and nonfatal MI exceeds 70% (11,15–17). Although stent thrombosis in patients on dual antiplatelet therapy typically occurs early and in the hospital, poor clinical outcomes, despite successful percutaneous restoration of flow, are common (8). Furthermore, six-month outcomes among survivors of stent thrombosis are not favorable, with mortality rates in the range of 20% to 25% (17). Mortality following stent thrombosis was particularly high (48%) in our registry. It is possible that this resulted from our characterizing unexplained death as presumptive stent thrombosis; some such deaths could have been due to arrhythmia, for example. Nonfatal MI occurred in 39% of the patients with stent thrombosis. However, these results are consistent with other studies in noting the significant morbidity and mortality associated with this increasingly rare complication of stent implantation.
First, subclinical stent thrombosis may have occurred in some patients and been totally undetected. It is also possible that over-diagnosis of stent thrombosis may have occurred due to the fact that we characterized all unexplained death as presumptive stent thrombosis, as discussed above. However, this methodology is common to most studies (randomized and observational) of stent thrombosis and is thought to reflect the incidence of the most clinical relevant outcomes of this complication (17,18). Second, in this study we have analyzed the frequency of stent thrombosis within 30 days; late stent thrombosis also occurs, and was not analyzed in this study. Again, this methodology is common to most studies of stent thrombosis, and late stent thrombosis (>30 days) is typically dealt with as a distinct clinical entity (19,20). Third, these data were collected by relying on patient interview, and therefore recall bias must be considered as a possible confounding factor. Finally, one patient was lost to follow-up; however, even if this patient had suffered stent thrombosis, the findings of this study would not change significantly.
Stent thrombosis in the first 30 days after stent placement and treatment with aspirin and a thienopyridine is rare in clinical practice, occurring in 0.5% of patients despite many clinical and angiographic risk factors previously associated with high frequencies of stent thrombosis. However, when stent thrombosis occurs, the majority of patients either die (48%) or have a nonfatal MI (39%).
☆ Dr. Berger has received grant support from, and served on a Scientific Advisory Board for, Sanofi and Bristol Myers Squibb, who make and co-market clopidogrel.
- accelerated clotting time
- adenosine diphosphate
- congestive heart failure
- confidence interval
- myocardial infarction
- odds ratio
- percutaneous coronary intervention
- Thrombolysis In Myocardial Infarction
- Received February 18, 2002.
- Revision received June 14, 2002.
- Accepted July 3, 2002.
- American College of Cardiology Foundation
- Hall P.,
- Nakamura S.,
- Maiello L.,
- et al.
- Urban P.,
- Macaya C.,
- Rupprecht H.J.,
- et al.
- Bertrand M.E.,
- Legrand V.,
- Boland J.,
- et al.
- Hasdai D.,
- Garratt K.N.,
- Holmes D.R. Jr.,
- Berger P.B.,
- Schwartz R.S.,
- Bell M.R.
- Schror K.
- Colombo A.,
- Hall P.,
- Nakamura S.,
- et al.
- Karrillon G.J.,
- Morice M.C.,
- Benveniste E.,
- et al.
- Moussa I.,
- Di Mario C.,
- Reimers B.,
- Akiyama T.,
- Tobis J.,
- Colombo A.
- Cutlip D.E.,
- Baim D.S.,
- Ho K.K.,
- et al.
- Berger P.B.,
- Bell M.R.,
- Rihal C.S.,
- et al.