Patients with atrial fibrillation (AF) who underwent percutaneous coronary intervention (PCI) are at elevated risk for bleeding and thromboembolic ischemic events. Currently, guidelines on antithrombotic treatment for these patients are based on weak consensus. We describe patterns and determinants of antithrombotic prescriptions in this population. The Antithrombotic Strategy Variability in Atrial Fibrillation and Obstructive Coronary Disease Revascularized with PCI Registry was an international observational study of 859 consecutive patients with AF who underwent PCI from 2009 to 2011. Patients were stratified by treatment at discharge with either dual antiplatelet therapy (DAPT; aspirin plus clopidogrel) or triple therapy (TT; warfarin plus DAPT). Bleeding and thromboembolism risks were assessed by the HAS-BLED and CHADS 2 scores, respectively, and predictors of TT prescription at discharge were identified. Major adverse cardiovascular events and clinically relevant bleeding (Bleeding Academic Research Consortium score ≥2) at 1-year follow-up were compared across antithrombotic regimens. Compared with patients on DAPT (n = 488; 57%), those given TT (n = 371; 43%) were older, with higher CHADS 2 scores, lower left ventricular ejection fraction, and more often had permanent AF, single-vessel coronary artery disease, and bare-metal stents. In multivariate analysis, increasing thromboembolic risk (CHADS 2 ) was associated with a higher rate of TT prescription at discharge (intermediate vs low CHADS 2 : odds ratio 2.2, 95% confidence interval [CI] 2.0 to 3.3, p <0.01; high vs low CHADS 2 : odds ratio 1.6, 95% CI 2.6 to 4.3, p <0.01 for TT). However, there was no significant association between bleeding risk and TT prescription in the overall cohort or within each CHADS 2 risk stratum. The rates of major adverse cardiovascular events were similar for patients discharged on TT or DAPT (20% vs 17%, adjusted hazard ratio 0.8, 95% CI 0.5 to 1.1, p = 0.19), whereas the rate of Bleeding Academic Research Consortium ≥2 bleeding was higher in patients discharged on TT (11.5% vs 6.4%, adjusted hazard ratio 1.8, 95% CI 1.1 to 2.9, p = 0.02). In conclusion, the choice of the intensity of antithrombotic therapy correlated more closely with the risk of ischemic rather than bleeding events in this cohort of patients with AF who underwent PCI.
Anticoagulation, conventionally with warfarin, is superior to dual antiplatelet therapy (DAPT) with aspirin plus a thienopyridine for prevention of ischemic stroke and systemic embolism in patients with atrial fibrillation (AF), whereas DAPT has proved superior to oral anticoagulation for prevention of stent thrombosis after percutaneous coronary intervention (PCI). The combination of anticoagulant and antiplatelet therapy is associated with a higher risk of bleeding than either strategy alone. For patients with AF who underwent PCI, optimum antithrombotic therapy is controversial, and clinical practice guidelines, based on consensus in the absence of high quality evidence from randomized trials, generally recommend individualized therapy that balances the risks of bleeding and thromboembolism for a given patient. Consensus documents favor triple therapy (TT) with warfarin plus DAPT for patients with AF who underwent PCI with stenting and CHADS 2 stroke risk scores >1. How these recommendations translate into clinical practice and how clinicians weigh ischemic and bleeding risks in the selection of antithrombotic therapy are poorly understood. We, therefore, examined a diverse patient cohort with AF who underwent PCI to determine the antithrombotic therapy prescribed in the context of stroke and bleeding risks and compare rates of ischemic and bleeding events in relation with antithrombotic therapy.
Methods
We identified 859 consecutive patients with nonvalvular AF who underwent successful PCI with stenting at 1 center in the United States (Mount Sinai Hospital, New York, New York) and 3 hospitals in Italy (Humanitas Clinical and Research Institute, Milan, Italy; San Raffaele Hospital, Milan, Italy; and Ferrarotto Hospital, Catania, Italy) from 2009 to 2011. To identify the study population, we queried databases from 4 centers for AF and PCI, and we selected 1,518 of 24,447 patients treated with PCI during the study period. Of them, 602 patients (40%) were excluded because of the presence of reversible causes of AF (thyrotoxicosis, pericarditis, etc.), valvular AF associated with mitral stenosis or prosthetic heart valves, or unsuccessful PCI, and 57 patients (4%) discharged on antithrombotic regimens other than DAPT or TT, including various combinations of antiplatelet monotherapy (aspirin, clopidogrel, or prasugrel) plus warfarin or a single agent. At each center, patients were treated according to local policies. Demographic, angiographic, and procedural data were collected from hospital charts according to prespecified definitions by trained research co-ordinators. The accuracy of the events and medications were validated using hospital and outpatient clinic records and mortality registries where appropriate. The respective institutional review boards approved the study, which involved collection of data at discharge and 1 year after PCI.
AF was classified as permanent if both the electrocardiogram at the time of PCI and an earlier tracing documented this arrhythmia and as paroxysmal or persistent if sinus rhythm was documented at the time of PCI, but an electrocardiogram showed AF in the previous 6 months. Bleeding and stroke risk assessments were based on the HAS-BLED and CHADS 2 scores, respectively. A HAS-BLED score >2 was considered high bleeding risk. CHADS 2 <2 was considered low risk, CHADS 2 = 2 intermediate risk, and CHADS 2 >2 high risk for stroke. DAPT was the combination of aspirin, 75 or 100 mg daily, plus a thienopyridine (clopidogrel, 75 mg daily, or ticlopidine, 250 twice daily) at discharge; TT was the combination of an oral vitamin K antagonist (warfarin or acenocoumarol) plus DAPT. Clinical outcomes evaluated 1 year after discharge included major adverse cardiac and cerebrovascular events (MACCEs) defined as all-cause mortality, nonfatal myocardial infarction or stroke, and bleeding, classified by the Bleeding Academic Research Consortium (BARC) criteria. Only bleeding requiring medical attention (BARC ≥2) qualified as an end point.
Statistical analyses were performed using STATA, version 12.1 (StataCorp, College Station, Texas). We excluded from the analysis patients with missing discharge therapy data (<1%), whereas other variables missing data (<5%) were not imputed, assuming that they were missing at random. Categorical data are presented as frequencies and were compared using Fisher’s exact test. Continuous data are presented as the mean ± 1 SD compared using factorial analysis of variance. Crude incidence rates (95% confidence interval [CI]) for MACCE and bleeding were determined using Kaplan-Meier curves; differences between groups were assessed using the log-rank test. Predictors of outcome were identified by multivariate analysis using Cox proportional hazards regression models. Independent predictors of TT prescriptions at discharge were assessed by logistic regression. Statistical significance was accepted at the 95% confidence level (p <0.05) without adjustment for multiple comparisons.
Results
Of 859 patients included, the mean age was 73 ± 9.6 years, 71% were men, and the mean CHADS 2 score was 2.7 ± 1.2. Slightly >1/2 of the patients (56.8%) were discharged on DAPT. Patients discharged on TT were older, more likely to present with left ventricular systolic dysfunction, higher CHADS 2 score, permanent AF, and single-vessel coronary artery disease, and were more often treated with bare-metal stents ( Table 1 ). Figure 1 shows patients stratified by CHADS 2 and HAS-BLED scores. Most had both high CHADS 2 and HAS-BLED scores (n = 403, 43.9%).
| Entire Cohort n= 859 | DAPT N=488 | TT N=371 | p | |
|---|---|---|---|---|
| Age (years) | 73±9.6 | 72±10 | 73±9.0 | 0.05 |
| Male | 71% | 70% | 75% | 0.10 |
| Caucasian | 77% | 78% | 76% | 0.57 |
| Diabetes mellitus | 41% | 38% | 43% | 0.20 |
| Hypertension | 94% | 92% | 95% | 0.10 |
| Chronic Kindney Disease | 62% | 63% | 63% | 0.76 |
| Left Ventricle Ejection Fraction ≤ 40% | 24% | 19% | 29% | <0.01 |
| Previous bleeding | 4.0% | 4.5% | 2.4% | 0.10 |
| Previous Carebrovascular Accident | 12% | 14% | 16% | 0.23 |
| H/o heart failure | 54% | 60% | 41% | 0.03 |
| CHADS 2 score | 2.7±1.2 | 2.5±1.2 | 2.9±1.1 | <0.01 |
| HAS-BLED score | 2.9±0.7 | 2.9±0.7 | 2.9±0.6 | 0.27 |
| Permanent Atrial Fibrillation | 57% | 42% | 71% | <0.01 |
| Clinical presentation | ||||
| Unstable Angina | 32% | 30% | 33% | 0.42 |
| Acute Miocardial Infarction | 24% | 27% | 21% | 0.05 |
| Angiographic | ||||
| Multi-vessel Coronary Artery Disease | 71% | 74% | 67% | 0.02 |
| Severe Calcification | 25% | 27% | 23% | 0.29 |
| % Diameter of Stenosis pre | 85±11 | 85±11 | 84±11 | 0.29 |
| % Diameter of Stenosis post | 6.9±8.4 | 6.3±7.6 | 7.6±8.3 | 0.02 |
| Stent length | 31±21 | 32±21 | 31±19 | 0.70 |
| Bare metal stent | 33% | 29% | 39% | <0.01 |
By univariate analysis, TT after PCI was more often prescribed for patients with high CHADS 2 scores (p trend <0.01, Figure 2 ) or high HAS-BLED scores (p <0.01, Figure 2 ). When adjusted for bleeding risk stratum (low [≤2] vs high [>2] HAS-BLED score), there was a significant trend for TT with increasing stroke risk (p trend <0.01), but there was no difference in TT use within each CHADS 2 stroke risk stratum (low, intermediate, and high) for patients grouped according to their risk of bleeding ( Figure 3 ).
Multivariate analysis revealed that a high CHADS 2 score (intermediate vs low: odds ratio [OR] 2.2, 95% CI 2.0 to 3.3, p <0.01; high vs low: OR 1.6, 95% CI 2.6 to 4.3, p <0.01) and male gender (OR 1.5, 95% CI 1.1 to 2.1, p = 0.02) were independently associated with TT prescriptions. Conversely, patients with nonpermanent (paroxysmal or persistent) AF were more often given DAPT at discharge (OR 0.25, 95% CI 0.33 to 0.45, p <0.01 for TT; Figure 4 ).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
