OUP user menu

Patients with peripheral arterial disease in the CHARISMA trial

Patrice P. Cacoub, Deepak L. Bhatt, P.Gabriel Steg, Eric J. Topol, Mark A. Creager
DOI: http://dx.doi.org/10.1093/eurheartj/ehn534 192-201 First published online: 9 January 2009

Abstract

Aims The aim of this study was to determine whether clopidogrel plus aspirin provides greater protection against major cardiovascular events than aspirin alone in patients with peripheral arterial disease (PAD).

Methods and results This is a post hoc analysis of the 3096 patients with symptomatic (2838) or asymptomatic (258) PAD from the CHARISMA trial. The rate of cardiovascular death, myocardial infarction (MI), or stroke (primary endpoint) was higher in patients with PAD than in those without PAD: 8.2% vs. 6.8% [hazard ratio (HR), 1.25; 95% CI 1.08, 1.44; P = 0.002]. Among the patients with PAD, the primary endpoint occurred in 7.6% in the clopidogrel plus aspirin group and 8.9% in the placebo plus aspirin group (HR, 0.85; 95% CI, 0.66–1.08; P = 0.18). In these patients, the rate of MI was lower in the dual antiplatelet arm than the aspirin alone arm: 2.3% vs. 3.7% (HR, 0.63; 95% CI, 0.42–0.96; P = 0.029), as was the rate of hospitalization for ischaemic events: 16.5% vs. 20.1% (HR, 0.81; 95% CI, 0.68–0.95; P = 0.011). The rates of severe, fatal, or moderate bleeding did not differ between the groups, whereas minor bleeding was increased with clopidogrel: 34.4% vs. 20.8% (odds ratio, 1.99; 95% CI, 1.69–2.34; P < 0.001).

Conclusion Dual therapy provided some benefit over aspirin alone in PAD patients for the rate of MI and the rate of hospitalization for ischaemic events, at the cost of an increase in minor bleeding.

Keywords
  • Peripheral vascular disease
  • Aspirin
  • Clopidogrel
  • Prognosis
See page 131 for the editorial comment on this article (doi:10.1093/eurheartj/ehn565)

Introduction

Peripheral arterial disease (PAD) is a common manifestation of atherosclerosis, affecting an estimated 27million people in Europe and North America.1 PAD is associated with an increased risk for cardiovascular events due to coexistence of coronary artery disease (CAD) and cerebrovascular disease (CVD), such events being more frequent than ischaemic limb events.2,3 There is a 20–60% increased risk for myocardial infarction (MI), a two- to six-fold increased risk of cardiovascular death, and a 40% increased risk of stroke in patients with PAD.4 These adverse consequences of PAD occur, in part, because many affected patients are not diagnosed, and those with known PAD often are undertreated.5,6

Antiplatelet therapy reduces the risk of MI, stroke, and vascular death in patients with PAD. The Antithrombotic Trialists’ Collaboration meta-analysis of 42 trials including 9717 patients with PAD found a 22% odds reduction for adverse cardiovascular events (MI, stroke, or vascular death) in patients with PAD treated with antiplatelet therapy compared with those who were not treated.7 Both aspirin and clopidogrel are recommended for patients with PAD.4

Among the 6452 patients with PAD in the CAPRIE trial, clopidogrel reduced the risk of MI, stroke, or vascular death by 23.8% more than aspirin.8 The combination of aspirin and clopidogrel is more effective in preventing adverse cardiovascular events than aspirin alone in high-risk groups including patients undergoing percutaneous coronary intervention,9,10 patients with acute coronary syndromes without ST-segment elevation,11 and patients with ST-elevation MI.12 It is not known, however, whether dual antiplatelet therapy with aspirin plus clopidogrel is more effective than a single antiplatelet agent in patients with PAD. Analyses derived from the CAPRIE and CURE studies suggested that the absolute benefit of clopidogrel over aspirin was amplified in certain high-risk subgroups of patients.13,14

The combination of aspirin plus clopidogrel was most recently evaluated in the CHARISMA trial, which enrolled patients with either established atherothrombotic disease or multiple risk factors for atherothrombotic events.15 In CHARISMA, dual antiplatelet therapy vs. aspirin alone was associated with a non-significant 7.1% relative risk reduction in MI, stroke, or cardiovascular death over a median of 28 months.16 In a post hoc subgroup analysis of the CHARISMA population, patients with documented prior MI, ischaemic stroke, or symptomatic PAD appeared to derive significant benefit from dual antiplatelet therapy.17

In this context, we hypothesized and pre-specified that long-term treatment with a combination of clopidogrel plus aspirin may provide greater protection against cardiovascular events than aspirin alone in the very high-risk subgroup of patients with either symptomatic or asymptomatic PAD from the CHARISMA trial.

Methods

The CHARISMA trial was a prospective, multicentre, randomized, double-blind, placebo-controlled study. It was approved by the Institutional Review Committee of each participating institution as well as by appropriate National Ethics Committees. All patients gave written informed consent. The details of the trial design have been published previously.18 Briefly, patients with documented CAD, documented CVD, documented symptomatic PAD, or multiple atherothrombotic risk factors were randomly assigned to receive clopidogrel (75 mg per day) plus low-dose aspirin (75–162 mg per day) or placebo plus low-dose aspirin and followed for a median of 28 months.

This study includes the 3096 patients with PAD who were identified in the CHARISMA study. Of these, 2838 (91.7%) were symptomatic and 258 (8.4%) were asymptomatic. To fulfil the symptomatic PAD inclusion criterion, patients had to have either current intermittent claudication together with an ankle-brachial index (ABI) ≤0.85, or a history of intermittent claudication together with a previous related intervention (amputation, surgical or catheter-based peripheral revascularization). Asymptomatic patients with an ABI < 0.90 were identified among those with multiple risk factors.

Endpoints

The primary efficacy endpoint was the first occurrence of MI, stroke (of any cause), or death from cardiovascular causes (including haemorrhage). The principal secondary efficacy endpoints were the first occurrence of MI, stroke, death from cardiovascular causes, hospitalization for unstable angina, a transient ischaemic attack, or a revascularization procedure (coronary, cerebral, or peripheral). The primary safety endpoint was severe bleeding according to the GUSTO definition,19 which includes fatal bleeding and intracranial haemorrhage, or bleeding that caused haemodynamic compromise requiring blood or fluid replacement, inotropic support, or surgical intervention. Moderate bleeding according to the GUSTO criteria,19 which includes bleeding that led to transfusion but did not meet the criteria for severe bleeding, and other bleeding events, which were qualified as minor, were also examined. The study clinical events committee, whose members were unaware of treatment assignments, adjudicated all primary endpoints.

Statistical analysis

Data were analysed on an intention-to-treat basis, as described previously for the entire CHARISMA cohort.15 All randomized subjects were included in the analysis. Subjects with no events were counted in the analysis and censored at the last known follow-up time (or time of death). Demographic and baseline characteristics were described using frequencies. Group comparisons were performed using the Pearson χ2 test. Cardiovascular endpoint rates were compared using hazard ratios (HRs) and 95% confidence intervals (CIs) generated from a Cox proportional hazard model and tested using the log-rank test. Odds ratios (ORs) and 95% CIs were generated for the bleeding endpoints. Statistical significance was accepted at the two-sided 0.05 level. Analyses were carried out using SAS® 8.02 (SAS Institute Inc., Cary, NC, USA). The proportional hazards assumption was examined by plotting the log(−log(survival)) vs. the log of survival time. The proportionality test in the SAS procedure for PROC PHREG was also implemented. Although there were some slight deviations from the normality assumption, the estimates provided should only be interpreted as the average effect over time. The authors had full access to the data and take responsibility for its integrity. All authors have read and agree to the manuscript as written.

Results

In the CHARISMA study, there were 15 603 participants, including 12 153 patients with established cardiovascular disease. Of these, 2838 patients had documented symptomatic PAD (Figure 1). In addition, there were 3284 patients with multiple atherothrombotic risk factors only. Of these, 258 patients had an ABI < 0.90 and were classified as patients with asymptomatic PAD. Hence, the current study includes the subset of 3096 patients with symptomatic or asymptomatic PAD.

Figure 1

Design of the study. PAD, peripheral arterial disease.

Profile of patients with peripheral arterial disease

Considering the entire CHARISMA cohort, patients with PAD (symptomatic or not) were older than patients without PAD (median age 66 vs. 64 years; P < 0.001, Table 1). A prior history of MI, stroke, transient ischaemic attack, and percutaneous coronary intervention was reported less often in patients with PAD than in patients without PAD. Inclusion criteria of the present study explain these seemingly contradictory results. The majority (9315/12 341) of the patients without PAD had established CAD or CVD by definition of the inclusion criteria, whereas 3026/12 341 had multiple risk factors only. By comparison, PAD patients only needed to have PAD to be included in this group. Carotid endarterectomy and peripheral angioplasty or bypass surgery was more frequent in the PAD group. Concerning the atherothrombotic risk factors, the group of patients with PAD had a greater prevalence of smokers and a lesser prevalence of hypercholesterolaemia, diabetes, and diabetic nephropathy than patients without PAD (though this had to do with the inclusion criteria for patients enrolled into the trial with multiple risk factors). Overall, 85.9% of the patients with PAD received cardiovascular drugs vs. 93.3% of the patients without PAD (P < 0.001), respectively, including 72.7% vs. 77.9% who received statins (P < 0.001), 12.2% vs. 14.6% who received other lipid-lowering agents (P = 0.001), and 36.0% vs. 43.1% received antidiabetic medications (P < 0.001). PAD patients tended to be on a lower dose of daily aspirin: daily aspirin < 100 mg 48.6% vs. 45.2%; daily aspirin = 100 mg 31.9% vs. 31.8%; daily aspirin > 100 mg 19.4% vs. 23.0% in PAD vs. no PAD, respectively (P < 0.001).

View this table:
Table 1

Baseline characteristics and concomitant medications of patients with or without peripheral arterial disease in the CHARISMA trial

Entire CHARISMA cohortP-value
PAD (n = 3096)No PAD (n = 12 341)
Demographics
 Age, median, years66 (59, 73)64 (56, 71)<0.001
 Female930 (30.0)3659 (29.6)0.671
 Race<0.001
  Caucasian2660 (85.9)9709 (78.7)
  Hispanic281 (9.1)1318 (10.7)
  Asian30 (1.0)739 (6.0)
  Black96 (3.1)379 (3.1)
  Other29 (0.9)196 (1.6)
Selected clinical characteristics
 Smoking status
  Current995 (32.1)2129 (17.3)<0.001
  Former1646 (53.2)5894 (47.8)<0.001
 Hypertension2237 (72.3)9109 (73.8)0.079
 Hypercholesterolaemia2160 (69.8)9239 (74.9)<0.001
 Congestive heart failure199 (6.4)717 (5.8)0.193
 Prior myocardial infarction772 (24.9)4591 (37.2)<0.001
 Atrial fibrillation124 (4.0)452 (3.7)0.369
 Prior stroke269 (8.7)3538 (28.7)<0.001
 Transient ischaemic attack217 (7.0)1627 (13.2)<0.001
 Diabetes1120 (36.2)5368 (43.5)<0.001
 Percutaneous coronary intervention462 (14.9)3045 (24.7)<0.001
 Coronary artery bypass graft574 (18.5)2480 (20.1)0.052
 Carotid endarterectomy290 (9.4)533 (4.3)<0.001
 Peripheral angioplasty or bypass1567 (50.6)167 (1.4)<0.001
 Diabetic nephropathy232 (7.5)1768 (14.3)<0.001
Concomitant medicationsa
 Aspirin3087 (99.7)12 303 (99.7)0.876
 Diuretics1495 (48.3)5846 (47.4)0.361
 Nitrates725 (23.4)2923 (23.7)0.754
 Calcium antagonists1263 (40.8)4420 (35.8)<0.001
 Beta-blockers1468 (47.4)7075 (57.3)<0.001
 Angiotensin II-receptor blockers734 (23.7)3229 (26.2)0.005
 Ramipril464 (15.0)2312 (18.7)<0.001
 Other angiotensin-converting-enzyme inhibitors1307 (42.2)5853 (47.4)<0.001
 Other antihypertensive agents445 (14.4)1467 (11.9)<0.001
 Statins2252 (72.7)9618 (77.9)<0.001
  Atorvastatin1070 (34.6)4453 (36.1)0.114
  Simvastatin1007 (32.5)4310 (34.9)0.012
  Pravastatin312 (10.1)1593 (12.9)<0.001
  Fluvastatin107 (3.5)383 (3.1)0.317
  Lovastatin102 (3.3)450 (3.6)0.346
  Other statins164 (5.3)754 (6.1)0.087
 Other lipid-lowering agents379 (12.2)1801 (14.6)0.001
  Fibrates222 (7.2)1092 (8.8)0.003
  Binding resins111 (3.6)534 (4.3)0.065
  Nicotinic acid91 (2.9)441 (3.6)0.084
 Antidiabetic medications1114 (36.0)5323 (43.1)<0.001
  Insulin544 (17.6)2127 (17.2)0.659
  Thiazolidinediones178 (5.7)1053 (8.5)<0.001
  Other oral hypoglycaemic agents837 (27.0)4472 (36.2)<0.001
  • PAD, peripheral arterial disease.

  • Data are presented as n (%) unless otherwise specified.

  • aValues indicate the maximal frequency of use of each agent at any time during the trial (assessed at baseline and at every follow-up visit).

Among the patients with PAD, asymptomatic patients were older than symptomatic patients by 2 years (median age 68 vs. 66 years; P = 0.029) and included a larger proportion of women (40.7% vs. 29.1%, P < 0.001) (Table 2). Asymptomatic patients had more frequent history of stroke and transient ischaemic attack, whereas prior MI and vascular interventions (apart from peripheral angioplasty or bypass) were reported with similar rates in the two groups. Atherothrombotic risk factors, other than smoking, were also more frequent in asymptomatic than symptomatic patients (though again, this was likely due to the inclusion criteria of the trial).

View this table:
Table 2

Baseline characteristics of patients with symptomatic or asymptomatic peripheral arterial disease

Patients with PADP-value
Symptomatic (n = 2838)Asymptomatic (n = 258)
Demographics
 Age, median, years66 (59, 73)68 (61, 74)0.029
 Female825 (29.1)105 (40.7)<0.001
 Race<0.001
  Caucasian2451 (86.4)209 (81.0)
  Hispanic263 (9.3)18 (7.0)
  Asian18 (0.6)12 (4.7)
  Black80 (2.8)16 (6.2)
  Other26 (0.9)3 (1.2)
Selected clinical characteristics
 Smoking status
  Current938 (33.1)57 (22.1)<0.001
  Former1529 (53.9)117 (45.3)0.009
 Hypertension2029 (71.5)208 (80.6)0.002
 Hypercholesterolaemia1948 (68.6)212 (82.2)<0.001
 Congestive heart failure173 (6.1)26 (10.1)0.013
 Prior myocardial infarction716 (25.2)56 (21.7)0.210
 Atrial fibrillation112 (3.9)12 (4.7)0.580
 Prior stroke237 (8.4)32 (12.4)0.027
 Transient ischaemic attack190 (6.7)27 (10.5)0.023
 Diabetes980 (34.5)140 (54.3)<0.001
 Percutaneous coronary intervention426 (15.0)36 (14.0)0.648
 Coronary artery bypass graft520 (18.3)54 (20.9)0.302
 Carotid endarterectomy267 (9.4)23 (8.9)0.795
 Peripheral angioplasty or bypass1553 (54.7)14 (5.4)<0.001
 Diabetic nephropathy186 (6.6)46 (17.8)<0.001
  • PAD, peripheral arterial disease.

  • Data are presented as n (%) unless otherwise specified.

Outcome of patients with peripheral arterial disease

The overall rate of cardiovascular death, MI, or stroke was 8.2% in patients with PAD and 6.8% in patients without PAD (HR, 1.25; 95% CI, 1.08–1.44; P = 0.002, Table 3). Patients with PAD had also higher rates of death from any cause (HR, 1.80; 95% CI, 1.54–2.11; P < 0.001), death from cardiovascular causes (HR, 1.73; 95% CI, 1.42–2.12; P < 0.001), MI (HR, 1.33, 95% CI, 1.05–1.68; P = 0.017), and hospitalization for ischaemic events (HR, 1.97; 95% CI, 1.78–2.17; P < 0.001). Though the rate of severe bleeding in patients with PAD (1.7%) was similar to that observed in patients without PAD (1.5%, P = 0.30), the rate of moderate bleeding was significantly higher in patients with PAD (2.2% vs. 1.6%, P = 0.032).

View this table:
Table 3

Composite and individual primary endpoints in patients with or without peripheral arterial disease in the CHARISMA trial

Entire CHARISMA cohortHR (95% CI)aP-value
PAD (n = 3096)No PAD (n = 12 341)
Efficacy endpoints
Cardiovascular death, MI, or stroke (primary endpoint)255 (8.2)843 (6.8)1.25 (1.08–1.44)0.002
Death from any cause221 (7.1)514 (4.2)1.80 (1.54–2.11)<0.001
Death from cardiovascular causes136 (4.4)327 (2.6)1.73 (1.42–2.12)<0.001
Myocardial infarctionb93 (3.0)290 (2.3)1.33 (1.05–1.68)0.017
Ischaemic strokeb71 (2.3)302 (2.4)0.97 (0.75–1.25)0.782
Strokeb82 (2.6)356 (2.9)0.94 (0.74–1.20)0.635
Hospitalizationc566 (18.3)1244 (10.1)1.97 (1.78–2.17)<0.001
Safety endpoints
Severe bleeding53 (1.7)180 (1.5)1.18 (0.86–1.60)0.301
Fatal bleeding13 (0.4)30 (0.2)1.73 (0.90–3.32)0.095
Primary intracranial haemorrhage9 (0.3)44 (0.4)0.81 (0.40–1.67)0.576
Moderate bleeding67 (2.2)198 (1.6)1.36 (1.03–1.79)0.032
Minor bleeding854 (27.6)3,227 (26.1)1.08 (0.98–1.17)0.105
  • CI, confidence interval; HR, hazard ratio; PAD, peripheral arterial disease.

  • Data are presented as n (%).

  • aFor the safety endpoints, the ORs and 95% CI are calculated.

  • bFatal plus non-fatal events.

  • cFor unstable angina, transient ischaemic attack, or revascularization.

Outcome of patients with symptomatic PAD was similar to that of asymptomatic PAD patients (data not shown), except for the rate of hospitalization for ischaemic events, which was higher in symptomatic patients (18.9% vs. 11.2%, P = 0.002).

Effect of clopidogrel or dual antiplatelet therapy in patients with peripheral arterial disease

In the subset of 3096 patients with PAD, 1545 patients were randomized to receive clopidogrel plus aspirin and 1551 patients were randomized to receive placebo plus aspirin. The baseline characteristics of these patients in the two randomized arms were well matched (Table 4). The only difference was the prevalence of patients with a history of hypertension, which was lower in the clopidogrel group than in the placebo group (69.9% vs. 74.6%, P = 0.004). Use of concomitant medication was also similar in the two treatment groups, except use of calcium antagonists, which was lower in the clopidogrel arm than in the placebo arm (38.5% vs. 43.1%, P = 0.010).

View this table:
Table 4

Baseline characteristics and concomitant medications of patients with peripheral arterial disease in the two treatment arms of the CHARISMA trial

Patients with PADP-value
Clopidogrel plus aspirin (n = 1545)Placebo plus aspirin (n = 1551)
Demographics
 Age, median, years66 (59, 73)66 (59, 73)0.734
 Female463 (30.0)467 (30.1)0.931
 Race0.478
  Caucasian1324 (85.7)1336 (86.1)
  Hispanic144 (9.3)137 (8.8)
  Asian14 (0.9)16 (1.0)
  Black52 (3.4)44 (2.8)
  Other11 (0.7)18 (1.2)
Selected clinical characteristics
 Smoking status
  Current483 (31.3)512 (33.0)0.297
  Former827 (53.5)819 (52.8)0.687
 Hypertension1080 (69.9)1157 (74.6)0.004
 Hypercholesterolaemia1069 (69.2)1091 (70.3)0.486
 Congestive heart failure106 (6.9)93 (6.0)0.327
 Prior myocardial infarction386 (25.0)386 (24.9)0.950
 Atrial fibrillation70 (4.5)54 (3.5)0.137
 Prior stroke128 (8.3)141 (9.1)0.426
 Transient ischaemic attack119 (7.7)98 (6.3)0.132
 Diabetes555 (35.9)565 (36.4)0.770
 Percutaneous coronary intervention228 (14.8)234 (15.1)0.797
 Coronary artery bypass graft282 (18.3)292 (18.8)0.681
 Carotid endarterectomy152 (9.8)138 (8.9)0.369
 Peripheral angioplasty or bypass794 (51.4)773 (49.8)0.388
 Diabetic nephropathy108 (7.0)124 (8.0)0.288
Concomitant medicationsa
 Aspirin1539 (99.6)1548 (99.8)0.342
 Diuretics745 (48.2)750 (48.4)0.940
 Nitrates339 (21.9)386 (24.9)0.053
 Calcium antagonists595 (38.5)668 (43.1)0.010
 Beta-blockers716 (46.3)752 (48.5)0.233
 Angiotensin II-receptor blockers374 (24.2)360 (23.2)0.515
 Ramipril227 (14.7)237 (15.3)0.647
 Other angiotensin-converting-enzyme inhibitors627 (40.6)680 (43.8)0.066
 Other antihypertensive agents220 (14.2)225 (14.5)0.832
 Statins1132 (73.3)1120 (72.2)0.509
  Atorvastatin523 (33.9)547 (35.3)0.407
  Simvastatin504 (32.6)503 (32.4)0.910
  Pravastatin159 (10.3)153 (9.9)0.693
  Fluvastatin62 (4.0)45 (2.9)0.090
  Lovastatin54 (3.5)48 (3.1)0.533
  Other statins80 (5.2)84 (5.4)0.768
 Other lipid-lowering agents174 (11.3)205 (13.2)0.097
  Fibrates105 (6.8)117 (7.5)0.420
  Binding resins49 (3.2)62 (4.0)0.217
  Nicotinic acid41 (2.7)50 (3.2)0.348
 Antidiabetic medications554 (35.9)560 (36.1)0.886
  Insulin274 (17.7)270 (17.4)0.811
  Thiazolidinediones86 (5.6)92 (5.9)0.662
  Other oral hypoglycaemic agents416 (26.9)421 (27.1)0.891
  • PAD, peripheral arterial disease.

  • Data are presented as n (%) unless otherwise specified.

  • aValues indicate the maximal frequency of use of each agent at any time during the trial (assessed at baseline and at every follow-up visit).

After a median duration of follow-up in PAD patients of 26 months (28 months in the entire CHARISMA cohort), the overall rate of cardiovascular death, MI, or stroke (primary efficacy endpoint) was 7.6% in the clopidogrel group and 8.9% in the placebo group (HR, 0.85; 95% CI, 0.66–1.08; P = 0.18, Table 5). A significant benefit was observed for the rate of MI, which was 2.3% in the clopidogrel group and 3.7% in the placebo group (HR, 0.63; 95% CI, 0.42–0.96; P = 0.029), and for the rate of hospitalization for ischaemic events, which was 16.5% in the clopidogrel group and 20.1% in the placebo group (HR, 0.81; 95% CI, 0.68–0.95; P = 0.011). The other efficacy outcomes were reported with similar rates in the two groups.

View this table:
Table 5

Composite and individual primary endpoints in patients with peripheral arterial disease in the two treatment arms of the CHARISMA trial

Patients with PADHR (95% CI)aP-value
Clopidogrel plus aspirin (n = 1545)Placebo plus aspirin (n = 1551)
Efficacy endpoints
 Primary endpoint117 (7.6)138 (8.9)0.85 (0.66–1.08)0.183
 Death from any cause104 (6.7)117 (7.5)0.89 (0.68–1.16)0.387
 Death from cardiovascular causes65 (4.2)71 (4.6)0.92 (0.65–1.28)0.613
 Myocardial infarctionb36 (2.3)57 (3.7)0.63 (0.42–0.96)0.028
 Ischaemic strokeb32 (2.1)39 (2.5)0.82 (0.52–1.32)0.416
 Strokeb36 (2.3)46 (3.0)0.79 (0.51–1.21)0.275
 Hospitalizationc255 (16.5)331 (20.1)0.81 (0.68–0.95)0.011
Safety endpoints
 Severe bleeding26 (1.7)27 (1.7)0.97 (0.56–1.66)0.901
 Fatal bleeding7 (0.5)6 (0.4)1.17 (0.39–3.49)0.776
 Primary intracranial haemorrhage3 (0.2)6 (0.4)0.50 (0.12–2.01)0.507
 Moderate bleeding38 (2.5)29 (1.9)1.32 (0.81–2.16)0.259
 Minor bleeding531 (34.4)323 (20.8)1.99 (1.69–2.34)<0.001
  • CI, confidence interval; HR, hazard ratio; PAD, peripheral arterial disease.

  • Data are presented as n (%).

  • aFor the safety endpoints, the ORs and 95% CI are calculated.

  • bFatal plus non-fatal events.

  • cFor unstable angina, transient ischaemic attack, or revascularization.

The rate of the primary safety endpoint (severe bleeding) was 1.7% in each treatment group (P = 0.90). The rate of minor bleeding was 34.4% in the patients treated with clopidogrel, when compared with 20.8% in the group treated with placebo (P < 0.001). Though the difference was not statistically significant, the rate of moderate bleeding was also higher in the clopidogrel group (2.5% vs. 1.9%, P = 0.26). Among patient requiring revascularization procedures, there was a trend for a benefit of clopidogrel plus aspirin combination in all subgroups (Table 6), although this was statistically significant only in the coronary stent group (3.1% vs. 4.4%; P = 0.05).

View this table:
Table 6

Vascular interventions during treatment in patients with peripheral arterial disease in the two treatment arms of the CHARISMA trial

InterventionPatients with PADP-value
Clopidogrel plus aspirin (n = 1545)Placebo plus aspirin (n = 1551)Total (n = 3096)
Peripheral arterial bypass surgery58 (3.8)79 (5.1)137 (4.4)0.070
Coronary stent48 (3.1)69 (4.4)117 (3.8)0.050
Percutaneous coronary angioplasty33 (2.1)49 (3.2)82 (2.6)0.076
Coronary artery bypass graft27 (1.7)34 (2.2)61 (2.0)0.374
Carotid endarterectomy16 (1.0)26 (1.7)42 (1.4)0.123
Leg amputationa12 (0.8)17 (1.1)29 (0.9)0.356
Carotid percutaneous angioplastyb11 (0.7)7 (0.5)18 (0.6)0.340
Other intervention for PAD142 (9.2)151 (9.7)293 (9.5)0.605
  • PAD, peripheral arterial disease.

  • Data are presented as n (%).

  • aFor critical leg ischaemia.

  • bWith or without stent.

Discussion

In this post hoc subgroup analysis of the CHARISMA trial, major cardiovascular events occurred more frequently in patients with PAD, either symptomatic or not, when compared with patients enrolled with CAD or CVD or with multiple risk factors. The rate of cardiovascular death, MI, or stroke (primary endpoint) was significantly higher in patients with PAD than in those without PAD. These results are consistent with those of the REACH registry, which found an event rate of 5.4% at 1 year for the same composite endpoint in patients with established PAD.20 Taken together, these observations underscore the fact that PAD has important prognostic implications for cardiovascular ischaemic events.2,3

The primary efficacy and safety endpoints occurred with similar frequency in PAD patients treated with the combination of clopidogrel and aspirin and in those treated with aspirin alone.

Nonetheless, this analysis suggests that in patients with PAD, dual antiplatelet therapy with clopidogrel plus aspirin may provide some benefits compared with aspirin alone. The combination of clopidogrel with aspirin therapy reduced the rate of MI and the rate of hospitalization for ischaemic events, but at the cost of an increased rate of minor bleeding. MI, the endpoint for which a benefit of dual therapy was found, was also the least frequent among the endpoints, both in the PAD subgroup and in the overall CHARISMA population. PAD patients, independent of the treatment group, showed an increased rate of moderate bleeding (2.2% vs. 1.6%; P = 0.032). The use of dual antiplatelet therapy to reduce cardiovascular events may be particularly relevant to high-risk groups of patients with atherothrombosis, as has been suggested previously.13,14,21 In a separate post hoc analysis of a ‘CAPRIE-like’ subgroup of patients with documented prior MI, ischaemic stroke, or symptomatic PAD from the CHARISMA trial, there was a significant 1.7% absolute risk reduction in the composite of cardiovascular death, MI, or stroke in patients treated with clopidogrel plus aspirin vs. aspirin alone.17 In that analysis, the PAD group included only symptomatic patients. We have extended the analyses to include all patients with PAD, whether symptomatic or asymptomatic, and specifically examined the efficacy of dual antiplatelet therapy in this population on the major secondary efficacy and safety endpoints. Our findings are consistent with observations made in the CAPRIE trial,8 which showed reduced rate of cardiovascular events with clopidogrel compared with aspirin in the PAD subgroup, though the number of PAD patients in CHARISMA was considerably less than in CAPRIE, thereby limiting statistical power. In CHARISMA, addition of clopidogrel to aspirin appeared to confer benefits in terms of reducing MI and hospitalizations among patients with PAD.

Given the well-known limitations and potential confounding of post hoc subset analyses,2224 our results should be viewed as merely hypothesis generating and need confirmation with appropriately designed prospective studies. Nevertheless, the PAD subgroup represented a large part (∼20%) of the entire CHARISMA cohort and included more than 3000 patients.

In conclusion, in the CHARISMA trial, patients with symptomatic or asymptomatic PAD showed a worse outcome than patients without PAD, i.e. CAD or CVD patients or those with multiple risk factors. Dual antiplatelet therapy with clopidogrel plus aspirin provided some benefit over aspirin alone in patients with symptomatic or asymptomatic PAD. Benefit was observed for the rate of MI and the rate of hospitalization for ischaemic events, at the cost of an increase in minor, though not moderate or severe bleeding. Such patients may benefit from antithrombotic therapy intensification beyond aspirin alone, a concept that future PAD trials will need to validate.

Conflict of interest: P.P.C has received research grants/honoraria/consultant fees from Astra Zeneca, Bristol–Myers Squibb, Encysive, Gilead, Roche, Sanofi Aventis, Schering Plough, and Servier. D.L.B. has received research grants/honoraria/consultant fees from Arena, Astra Zeneca, Bayer, Bristol–Myers Squibb, Cardax, Centocor, Cogentus, Daiichi-Sankyo, Eisai, Eli Lilly, Ethicon, Glaxo SmithKline, Heartscape, Johnson & Johnson, McNeil, Medtronic, Millenium, Otsuka, Paringenix, PDL, Portola, Sanofi Aventis, Schering Plough, Scios, The Medicines Company, tns Healthcare, and Vertex; he also provided expert testimony regarding anti-thrombotic therapy (the compensation was donated to a non-profit organization; >2 years). P.G.S. has received research grants/honoraria/consultant fees from Astra Zeneca, Boehringer-Ingelheim, Bristol–Myers Squibb, Glaxo SmithKline, Merck Sharp and Dohme-Chibret, Nycomed, Sanofi Aventis, Servier, Takeda, The Medicines Company, and ZLB-Behring. E.J.T. has received research grants from Bristol–Myers Squibb and Sanofi Aventis. M.A.C. has received research grants/honoraria/consultant fees from Bristol–Myers Squibb, Genzyme, Sanofi Aventis, and Sigma Tau.

Funding

The CHARISMA trial was funded by Sanofi-Aventis and Bristol–Myers Squibb.

References

View Abstract