OUP user menu

★ editor's choice ★

Combining warfarin and antiplatelet therapy after coronary stenting in the Global Registry of Acute Coronary Events: is it safe and effective to use just one antiplatelet agent?

Michael C. Nguyen, Yean L. Lim, Antony Walton, Jeffrey Lefkovits, Giancarlo Agnelli, Shaun G. Goodman, Andrzej Budaj, Dietrich C. Gulba, Jeanna Allegrone, David Brieger
DOI: http://dx.doi.org/10.1093/eurheartj/ehm186 1717-1722 First published online: 11 June 2007

Abstract

Aims To identify factors associated with the use of single or dual antiplatelet therapy in patients prescribed warfarin following coronary stenting and to investigate whether single (aspirin or thienopyridine) vs. dual antiplatelet therapy plus warfarin leads to an excess of adverse outcomes.

Methods and results We analysed data from 800 patients with an acute coronary syndrome who underwent coronary stenting (130 patients received a drug-eluting stent) and were discharged on warfarin and either dual (n = 580) or single (n = 220) antiplatelet therapy. The use of single antiplatelet therapy was more common in Europe than in the USA (34 vs. 17%, P < 0.001). There was no difference in major bleeding in hospital or in 6-month mortality or myocardial infarction. In the single antiplatelet group, the use of either aspirin or thienopyridine (clopidogrel or ticlopidine) in combination with warfarin resulted in similar outcomes.

Conclusion Use of single vs. dual antiplatelet therapy and warfarin following stenting is common. In this observational study, there was no difference in mortality or myocardial infarction at 6 months; however, larger trials are needed to assert any firm recommendations.

  • Warfarin
  • Antiplatelet therapy
  • Acute coronary syndromes

Introduction

Dual antiplatelet therapy with aspirin and a thienopyridine following coronary stenting is superior to aspirin alone in reducing cardiovascular events in both the acute coronary syndrome (ACS) and the elective setting;1,2 however, dual antiplatelet therapy is associated with an increased risk of bleeding.3 Dual antiplatelet therapy tends to be prescribed long term, and the duration of treatment now often extends to 12 months or longer.2 Dual antiplatelet therapy has also been proven to be superior in terms of safety and efficacy when compared with anticoagulation with warfarin following coronary stenting.46

Identifying the optimal regimen for antiplatelet therapy in patients requiring warfarin following coronary stenting is an area that has been understudied. Small, single-centre studies79 and larger observational studies10 and meta-analyses11 have examined the safety of antiplatelet therapy in combination with warfarin. These studies primarily analysed safety rather than efficacy outcomes and do not address the comparison of single vs. dual antiplatelet in combination with warfarin following coronary stenting in patients with a strong indication for warfarin.

Clinical trials in patients with ACS have supported an early invasive approach in those at high risk of a subsequent coronary event.1214 These patients are triaged to coronary angiography, often leading to percutaneous coronary revascularization. They tend to comprise an elderly population with often substantial vascular, valvular, and conduction system disease. Therefore, a significant proportion of these patients also requires warfarin for various co-morbid conditions (e.g. atrial fibrillation, mechanical valve replacement, deep-vein thrombosis/pulmonary embolus, or large anterior myocardial infarcts).

In patients requiring warfarin, standard practice is to combine both antiplatelet and antithrombotic drugs; however, the use of single or dual antiplatelet in combination with warfarin for this population needs further investigation. The choice of single antiplatelet (aspirin vs. thienopyridine) also needs to be examined. In this study, we attempted to identify factors associated with the use of single or dual antiplatelet therapy in patients who required warfarin following coronary stenting in an unselected population presenting with an ACS. We compared the clinical outcomes of each regimen, including those in patients prescribed either aspirin or a thienopyridine (either clopidogrel or ticlopidine) in combination with warfarin.

Methods

Full details of the GRACE methods have been published.1517 GRACE is designed to reflect an unbiased population of patients with ACS, irrespective of geographical region. A total of 113 hospitals located in 14 countries in North and South America, Europe, Australia, and New Zealand have contributed data to this observational study.

Patients entered in the registry had to be at least 18 years old and alive at the time of hospital presentation, be admitted for ACS as a presumptive diagnosis (i.e. have symptoms consistent with acute ischaemia), and should have at least one of the following: electrocardiographic changes consistent with ACS, serial increases in serum biochemical markers of cardiac necrosis, and/or documentation of coronary artery disease. The qualifying ACS must not have been precipitated by significant non-cardiovascular co-morbidity (e.g. trauma or surgery). Approximately 6 months after hospital discharge, patients were followed-up to ascertain the occurrence of selected long-term study outcomes. Where required, study investigators received approval from their local hospital ethics or institutional review board.

The study aimed to enrol an unbiased population and sites were encouraged to recruit the first 10 to 20 consecutive eligible patients each month. Data were collected by trained coordinators using standardized case report forms. Demographic characteristics, medical history, presenting symptoms, duration of pre-hospital delay, biochemical and electrocardiographic findings, treatment practices, and a variety of hospital outcome data were collected. Standardized definitions of all patient-related variables, clinical diagnoses, and hospital complications and outcomes were used.15

We analysed data from 800 patients (entered between April 1999 and September 2006) who underwent coronary stenting following presentation with an ACS and who were subsequently discharged on warfarin and dual antiplatelet therapy or warfarin and single antiplatelet therapy. Baseline demographics, geographical differences, hospital interventions, hospital events, and 6-month outcomes were analysed.

Statistical analysis

Data are expressed as percentages, or as medians and interquartile ranges, as appropriate. Differences between warfarin/single and warfarin/dual antiplatelet groups were evaluated by χ2 or Fisher's exact test for categorical variables and the Wilcoxon rank-sum test for continuous variables. Two-sided univariate analyses were performed.

Results

Of the patients in this cohort, 580 (73%) received warfarin and dual antiplatelet therapy and 220 (28%) warfarin and single antiplatelet therapy. The use of single antiplatelet therapy was more frequent outside the USA (35 vs. 17%, P < 0.001; Figure 1). Within the single antiplatelet group, 107 patients received aspirin and 113 a thienopyridine. The use of a thienopyridine (clopidogrel or ticlopidine) in combination with warfarin was most common in Australia/New Zealand when compared with the rest of the world (97 vs. 43%, P < 0.001).

Figure 1

Geographical variation in combination regimen at discharge. Four-way P-value less than 0.001.

The baseline characteristics of patients receiving warfarin/dual antiplatelet therapy and warfarin/single antiplatelet therapy are shown in Table 1. Patients receiving single antiplatelet therapy were older and more likely to have a history of prior angina.

View this table:
Table 1

Patients' demographic and baseline characteristics and clinical presentation according to antiplatelet therapy at discharge

Combination discharge therapy
Warfarin/dual antiplatelet (n = 580)Warfarin/single antiplatelet (n = 220)P-value
Demographics n (%)
 Median age, years (IQR)55–75 (64)58–77 (66)0.02
 Men432 (74)129 (70)0.23
 Prior angina227 (39)105 (48)0.02
 Prior myocardial infarction59 (27)58 (26)0.78
 Prior heart failure60 (10)29 (13)0.26
 Prior coronary intervention108 (19)34 (16)0.32
 Prior CABG surgery86 (15)27 (12)0.36
 Prosthetic valve20/356 (5.6)5/124 (4.0)0.49
 Smoker (current or former)336 (58)116 (53)0.16
 Diabetes130 (23)49 (23)0.99
 Hypertension331 (57)129 (59)0.73
 Hyperlipidaemia301 (52)100 (47)0.16
 Atrial fibrillation130 (22)52 (24)0.67
 Major surgery/trauma26 (4.5)13 (5.9)0.41
Clinical presentation n (%)
 Cardiac arrest15 (2.6)8 (3.7)0.41
 Killip class I452 (80)180 (84)0.34
 Killip class II–IV114 (19)35 (16)
 STEMI355 (61)134 (61)0.97
 Non-STEMI134 (23)50 (23)
 Unstable angina91 (16)36 (16)
  • CABG, coronary artery bypass grafting; IQR, interquartile range; STEMI, ST-segment elevation myocardial infarction.

A total of 226 (28%) patients were taking warfarin before presentation; 182 had a history of atrial fibrillation. The specific indications for warfarin therapy during admission were not collected on the GRACE case report form. However, of the 574 (72%) patients not on warfarin therapy before hospitalization, the assumed indications for in-hospital initiation of long-term anticoagulation are indicated in Table 2 and included new onset atrial fibrillation, anterior myocardial infarcts, valve surgery, and venous thrombo-embolism. In total, 299 (37%) of the cohort had either a history of atrial fibrillation or developed atrial fibrillation during admission.

View this table:
Table 2

Assumed indications for warfarin therapy

Indication n (%)Prior warfarin (n = 226)New warfarin therapy (n = 574)
Atrial fibrillation or flutter182 (80)137 (24)
STEMI0343 (60)
Prosthetic valve surgery20 (9)0
Venous thrombo-embolism20 (9)12 (2)
Unidentified4 (2)82 (14)
  • STEMI, ST-segment elevation myocardial infarction.

Patients discharged on warfarin/single antiplatelet therapy were less likely than those given warfarin/dual therapy to receive unfractionated heparin, glycoprotein IIb/IIIa antagonists, beta-blockers, or statins while in hospital; the use of low-molecular-weight heparin did not differ between groups. They were also less likely to have received a drug-eluting stent, although 28 (22%) of the patients who received a drug-eluting stent were discharged on single antiplatelet therapy (Table 3).

View this table:
Table 3

Prior medications, and hospital treatment and interventions, according to combination regimen

Combination discharge therapy
Warfarin/dual antiplatelet (n = 580)Warfarin/single antiplatelet (n = 220)P-value
Prior medications n (%)
 Aspirin179 (31)61 (28)0.39
 Warfarin167 (29)59 (27)0.57
 Thienopyridines31 (5)15 (7)0.44
Hospital treatment and interventions n (%)
 Aspirin574 (99)198 (90)<0.001
 Thienopyridines408 (95)108 (68)<0.001
 Unfractionated heparin433 (75)136 (62)<0.001
 LMWH307 (53)118 (54)0.87
 GP IIb/IIIa inhibitors356 (61)108 (49)0.003
 Beta-blockers532 (92)187 (85)0.008
 ACE-inhibitors479 (83)174 (79)0.22
 Fibrinolytic drug102 (18)33 (15)0.37
 Pulmonary artery catheter34 (6)22 (10)0.04
 Drug-eluting stent102/358 (28)28/124 (22)a
  • ACE, angiotensin-converting enzyme; GP, glycoprotein; LMWH, low-molecular-weight heparin.

  • aData on the type of stent (drug-eluting vs. bare-metal stent) were collected from 482 patients.

Patients discharged on warfarin/single antiplatelet therapy were more likely to have experienced atrial fibrillation or congestive heart failure during their admission, but there were no differences in the incidences of myocardial infarction, recurrent ischaemia, cardiogenic shock, or major bleeding (Table 4).

View this table:
Table 4

Hospital events according to combination regimen

Event n (%)Combination discharge therapy
Warfarin/dual antiplatelet (n = 580)Warfarin/single antiplatelet (n = 220)P-value
Cardiogenic shock33 (5.7)17 (7.7)0.27
Myocardial infarction > 24 h/re-infarction48 (8.3)26 (12)0.12
Recurrent ischaemic symptoms164 (28)57 (26)0.49
Atrial fibrillation/flutter127 (22)66 (30)0.01
Stroke6 (1.0)7 (3.2)0.05
Congestive heart failure128 (22)65 (30)0.02
Major bleeding34 (5.9)10 (4.6)0.46

The use of antiplatelet therapy at 6 months differed significantly between the two groups with those discharged on single antiplatelet being more likely to be on no antiplatelet at 6 months (Table 5).

View this table:
Table 5

Antiplatelet use at 6-month follow-up (671 of 800 patients had completed medication at 6-month follow-up)

Antiplatelet use n (%)Combination discharge therapy
Warfarin/dual antiplatelet (n = 479)Warfarin/single antiplatelet (n = 192)P-value
Dual antiplatelet at follow-up116 (24)24 (12)
Single antiplatelet at follow-up235 (49)94 (49)<0.001
None at follow-up128 (27)74 (39)

There was no difference in rates of 6-month mortality or myocardial infarction between the groups. The frequency of 6-month stroke was lower in the warfarin/dual antiplatelet group (0.7 vs. 3.4%, P = 0.02), although the number of events was small (Table 6).

View this table:
Table 6

Six-month outcomes in the overall cohort, in patients with atrial fibrillation, and according to use of aspirin or thienopyridine

Warfarin/dual antiplateletWarfarin/single antiplateletP-value
Overall cohort n (%)
 Death23/453 (5.1)12/184 (6.5)0.47
 Stroke3/426 (0.7)6/179 (3.4)0.02a
 Unscheduled PCI45/424 (10.6)22/176 (12.5)0.50
 Myocardial infarction13/391 (3.3)7/154 (4.5)0.49
Cohort with atrial fibrillation n (%)
 Death9/156 (5.8)7/75 (9.3)0.32
 Stroke0/148 (0)1/71 (1.4)0.32
 Unscheduled PCI13/148 (8.8)7/68 (10.3)0.72
 Myocardial infarction3/138 (2.2)2/61 (3.3)0.64
Aspirin vs. thienopyridine n (%)Warfarin/aspirin (n = 107)Warfarin/thienopyridine (n = 113)
 Death7/91 (7.7)5/93 (5.4)0.52
 Stroke4/88 (4.5)2/91 (2.2)0.44
 Myocardial infarction3/75 (4.0)4/79 (5.1)1.00
 Unscheduled PCI15/87 (17.2)7/89 (7.9)0.06
  • PCI, percutaneous coronary intervention.

  • aFisher's exact test.

Among patients with atrial fibrillation, there were no significant differences in 6-month outcomes between the single and dual antiplatelet groups (Table 6). Of the patients discharged on warfarin/single antiplatelet therapy, 107 (49%) received aspirin and 113 (51%) a thienopyridine. The aspirin/warfarin combination was associated with similar 6-month outcomes to that of thienopyridine/warfarin (Table 6).

Of the 130 patients discharged on warfarin following placement of a drug-eluting stent, antiplatelet therapy was more likely to be continued for 6 months. There were no differences in outcomes between patients discharged on single vs. dual antiplatelet therapy (Table 7).

View this table:
Table 7

Antiplatelet use at 6-months, and 6-month outcomes, among the cohort who received a drug-eluting stent

Combination discharge therapy
Warfarin/dual antiplatelet (n = 75)Warfarin/single antiplatelet (n = 25)P-value
Antiplatelet use at follow-up n (%)a
 Dual antiplatelet44 (59)8 (32)0.02
 Single antiplatelet14 (19)11 (44)
 No antiplatelet17 (22)6 (24)
Six-month outcomes n (%)(n = 102)(n = 28)
 Death5/67 (7.5)3/25 (12)0.67
 Stroke0/62 (0)0/21 (0)
 Unscheduled PCI14/62 (23)3/22 (14)0.54
 Myocardial infarction2/63 (3.2)0/23 (0)1.0
  • PCI, percutaneous coronary intervention.

  • aInformation on antiplatelet therapy at 6-month follow-up was available in only 100 of the 130 patients.

Discussion

The aim of this study is to review an area of clinical practice that is commonly encountered but grossly understudied. Our real-world data suggest that, in patients with an ACS who are discharged on antiplatelet therapy and warfarin following coronary stenting, the use of single vs. dual antiplatelet therapy leads to similar 6-month efficacy outcomes. A dual antiplatelet regimen of aspirin and a thienopyridine is the optimal treatment strategy following coronary stenting based on superior safety and efficacy when compared with aspirin alone or aspirin in combination with warfarin.5,6 In the Stent Anticoagulation Restenosis Study (STARS),5 three antiplatelet/antithrombotic regimens (aspirin alone, aspirin/ticlopidine, and aspirin/warfarin) were compared in an elective coronary stenting cohort. The Full Anticoagulation vs. Aspirin and Ticlopidine (FANTASTIC) trial6 compared a regimen of dual antiplatelet therapy vs. aspirin/warfarin in both elective and unplanned (to salvage failed angioplasty or optimize the results of balloon angioplasty) cohorts. Both randomized trials demonstrated a significant benefit in terms of safety and efficacy in favour of dual antiplatelet therapy alone. The STARS trial5 indicated that most of the benefits were from reductions in subacute stent thrombosis in the dual antiplatelet group. Both trials occurred during the early bare-metal stent era and involved a cohort of patients primarily undergoing elective procedures.

No large trial has addressed the population of patients with a strong indication for warfarin or presenting with an ACS. Our registry analysis, with the largest sample size in the current literature, reflects real-world practice in a real-world population, investigating a common clinical problem that clinicians face on a daily basis. The fine balance between safety and efficacy in this cohort needs careful consideration. Arab et al.18 performed a systematic review of the literature on the optimal antiplatelet/antithrombotic regimen in those requiring long-term anticoagulation who undergo coronary stenting and found no significant studies or randomized trials addressing this issue.

Our results demonstrate varying practices with significant differences between the USA and other parts of the world such as Europe, Australasia, and Argentina/Brazil. The type of single antiplatelet chosen in combination with warfarin also varied significantly, with Australasia almost exclusively using a thienopyridine in combination with warfarin. The regional variations may reflect differences in healthcare systems as well as economic and social influences.

Our results suggest that, in combination with warfarin, the use of single antiplatelet therapy may be a safe treatment option in selected patients, with similar 6-month efficacy outcomes when compared with dual antiplatelet therapy. There was no difference in major bleeding in-hospital between the groups; however, 6-month bleeding outcomes were not captured. There was a trend towards an increase in events in patients receiving single antiplatelet therapy at 6 months; however, the only significant difference in favour of dual antiplatelet therapy involved stroke at 6 months (0.7 vs. 3.4%, P = 0.02). This result should be interpreted with caution as the event rates were extremely small (3 vs. 6). The type of stroke (ischaemic vs. haemorrhagic) was not identified in this registry.

The use of either aspirin or a thienopyridine as the single antiplatelet in combination with warfarin also proved to be equivalent in terms of efficacy. Thienopyridine use was significantly reduced in patients with a history of major surgery or trauma (1.8 vs. 10.3%, P < 0.01). The incidence of subacute and late stent thrombosis was not collected, but it is a rare event that would be captured in the 6-month myocardial infarction and mortality outcome data. Most patients had ceased dual antiplatelet therapy at 6 months (reflecting a cohort receiving mainly bare-metal stents). There was a significant difference between the two groups, with those discharged on single antiplatelet therapy more likely to be on no antiplatelet at 6 months (39 vs. 27%, P < 0.001) (Table 6). Despite this, there was no excess in events. In the single antiplatelet group, there were only seven myocardial infarcts at 6 months when compared with 13 in the dual antiplatelet group (4.5 vs. 3.3%, P = 0.49) and 12 deaths compared with 23 (6.5 vs. 5.1%, P = 0.47).

The perceived increased risk of late stent thrombosis in those receiving drug-eluting stents has resulted in a recommendation of prolonged dual antiplatelet therapy in these patients.19,20 An alternative strategy of bare-metal stenting in those requiring warfarin may avoid the need for long-term dual antiplatelet therapy. In this registry, which has been accruing data since 1999, only a minority of patients received drug-eluting stents. Among the warfarin/dual antiplatelet group, 102 of 358 (28%) patients received drug-eluting stents when compared with 28 of 124 (22%) in the warfarin/single antiplatelet. Analysis of this small cohort of patients revealed no difference in 6-month outcomes between the two groups (Table 7). However, as this registry is ongoing, we will have the opportunity to describe practice and associated outcomes among these patients in the future.

The optimal antiplatelet strategy for stented patients requiring warfarin is an unresolved clinical question. Practice at present is guided by the clinician's discretion, with no significant evidence to date to validate any one regimen. Although ours is an observational study with a limited number of events, the data suggest the use of single antiplatelet therapy combined with warfarin in patients with an indication for long-term anticoagulation to be an acceptable management option. There remains a pressing need for further investigation into this important area.

Strengths and limitations

GRACE is a large, ongoing, multinational registry. The strength of this study stems from its assessment of a real-world population. It highlights the varying practices from region to region, reflecting both economical and social differences but also the fact that there is no clear evidence or guidelines in this cohort of patients. The GRACE registry is the largest multinational registry to include the spectrum of ACS patients and follows standardized criteria for defining ACS and hospital outcomes allowing an accurate and reliable data set. The limitations of our registry are similar to those of any observational study. The lack of substantial numbers (even though the study is significantly larger than any previous study in this area) also makes it difficult to draw firm conclusions. The study does not capture data on long-term bleeding outcomes, which is important to evaluate the ongoing safety of each regimen.

Acknowledgements

The authors thank the physicians and nurses participating in GRACE. Further information about the project, along with the complete list of participants, can be found at www.outcomes.org/grace. The authors are grateful to Sophie Rushton-Smith, who provided editorial support and was funded by sanofi-aventis. GRACE is supported by an unrestricted educational grant from sanofi-aventis, Paris, France, to the Center for Outcomes Research, University of Massachusetts Medical School, Worcester, MA, USA. Sanofi-aventis had no involvement in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

Conflict of interest: The following authors have no conflict of interest: M.C.N., Y.L.L., A.W., J.L., and J.A. The following authors have conflict of interest: G.A., consultant and/or speaker for sanofi-aventis; S.G.G., research grant support and/or consultant and/or speaker for sanofi-aventis and Bristol-Myers Squibb; A.B., consultant and/or speaker for sanofi-aventis; D.C.G., consultant and/or speaker for sanofi-aventis; D.B., consultant and/or speaker for sanofi-aventis; S.R.-S., funded by sanofi-aventis.

Footnotes

  • Preliminary results of this study were presented as an Abstract at the 2006 Annual Scientific Session of the American College of Cardiology, Atlanta, GA, USA and were published in J Am Coll Cardiol 2006;47:252A.

References

View Abstract