Bleeding Complications With Dual Antiplatelet Therapy Among Patients With Stable Vascular Disease or Risk Factors for Vascular Disease
Results From the Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) Trial
Background— Uncertainty exists about the frequency, correlates, and clinical significance of bleeding with dual antiplatelet therapy (DAPT), particularly over an extended period in a stable population. We sought to determine the frequency and time course of bleeding with DAPT in patients with established vascular disease or risk factors only; identify correlates of bleeding; and determine whether bleeding is associated with mortality.
Methods and Results— We analyzed 15 603 patients enrolled in the Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trial, a double-blind, placebo-controlled, randomized trial comparing long-term clopidogrel 75 mg/d versus placebo; all patients received aspirin (75 to 162 mg) daily. Patients had either established stable vascular disease or multiple risk factors for vascular disease without established disease. Median follow-up was 28 months. Bleeding was assessed with the use of the Global Utilization of Streptokinase and t-PA for Occluded Coronary Arteries (GUSTO) criteria. Severe bleeding occurred in 1.7% of the clopidogrel group versus 1.3% on placebo (P=0.087); moderate bleeding occurred in 2.1% versus 1.3%, respectively (P<0.001). The risk of bleeding was greatest the first year. Patients without moderate or severe bleeding during the first year were no more likely than placebo-treated patients to have bleeding thereafter. The frequency of bleeding was similar in patients with established disease and risk factors only. In multivariable analysis, the relationship between moderate bleeding and all-cause mortality was strong (hazard ratio, 2.55; 95% confidence interval, 1.71 to 3.80; P<0.0001), along with myocardial infarction (hazard ratio, 2.92; 95% confidence interval, 2.04 to 4.18; P<0.0001) and stroke (hazard ratio, 4.20; 95% confidence interval, 3.05 to 5.77; P<0.0001).
Conclusions— In CHARISMA, there was an increased risk of bleeding with long-term clopidogrel. The incremental risk of bleeding was greatest in the first year and similar thereafter. Moderate bleeding was strongly associated with mortality.
Clinical Trial Registration— URL: http://www.clinicaltrials.gov. Unique identifier: NCT00050817.
Received July 20, 2009; accepted April 15, 2010.
Dual antiplatelet therapy (DAPT) with aspirin and clopidogrel reduces thrombotic events among patients with ST-segment and non–ST-segment elevation acute coronary syndromes and undergoing percutaneous coronary interventions with both bare and drug-eluting stents.1–5 However, because of bleeding risk, uncertainty remains about the optimal duration of clopidogrel therapy in these populations.
Clinical Perspective on p 2583
Analyses of bleeding risk among high-risk vascular patients taking antiplatelet therapy for up to 1 year have been reported, and risk factors have been identified.3,4 However, little is known about risks and correlates of bleeding in stable patients over a long duration. In the Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trial, aspirin and either clopidogrel or placebo were administered for a median of 28 months.6,7 Clopidogrel did not reduce thrombotic events in the overall study population. However, there appeared to be heterogeneity between patients with stable vascular disease, in whom clopidogrel led to a 12% relative decrease in cardiovascular death, myocardial infarction (MI), or stroke versus those with risk factors only, in whom there was a 20% increase in these events.7 This raises the issue of whether bleeding accounts for increased risk in patients with risk factors only and whether bleeding risk is sufficiently high to argue against DAPT in stable patients with vascular disease.
Accordingly, we performed this analysis of CHARISMA to determine the frequency and time course of bleeding with DAPT in patients with and without established vascular disease. We also sought to determine correlates of bleeding and its association with mortality.
Study Design and Participants
The design, methods, and primary results of CHARISMA have been published.6,7 Briefly, CHARISMA was a multicenter, double-blind, randomized, placebo-controlled trial comparing 75 mg/d of clopidogrel versus placebo in patients aged ≥45 years either with documented vascular disease or at high risk of having disease. All patients received aspirin 75 to 162 mg daily, by physician preference. Patients were excluded if they had proven indications for clopidogrel (recent acute coronary syndrome or stent). Patients were also excluded if they were believed to be at particularly high risk of bleeding (severe hepatic insufficiency, current peptic ulceration, proliferative diabetic retinopathy), had previous severe bleeding (gastrointestinal, urological, intraocular, or intracranial), had a bleeding diathesis or coagulopathy, or were receiving dipyridamole. Patients requiring warfarin were excluded from the trial. If warfarin treatment became necessary after enrollment, the protocol stated that “the study drug should be temporarily interrupted (if possible, for less than 4 weeks). In any case, investigators are strongly encouraged to restart the study drug as soon as possible after such concomitant therapy is stopped.” At some time during the trial, 11% of patients received an antiplatelet agent, and 12% received an anticoagulant (167 in the clopidogrel group, 178 in the placebo group). Clinical follow-up was required at 1, 3, and 6 months and every 6 months thereafter until study end (median, 28 months after enrollment).
End Points Analyzed
The primary safety end point was severe bleeding as defined by the Global Utilization of Streptokinase and t-PA for Occluded Coronary Arteries (GUSTO) criteria, which includes fatal bleeding and intracranial hemorrhage, or bleeding causing hemodynamic compromise requiring blood or fluid replacement, inotropes, or surgery.8 The frequencies of moderate bleeding, defined as bleeding requiring transfusion not characterized as severe, and moderate/severe and severe bleeding were also analyzed. The primary efficacy end point was the first occurrence of cardiovascular death/MI/stroke of any cause. Death due to hemorrhage was considered cardiovascular. All components of the primary efficacy and safety end points and moderate bleeding were adjudicated by a Clinical Events Committee blinded to treatment assignment. Follow-up for the primary efficacy end point was complete in 99.6% of patients.
Analyses were based on intention to treat. Hypothesis tests were performed with the use of 2-sided tests at 5% significance level. Baseline characteristics were compared with χ2 tests for discrete variables and t tests for continuous variables. Event rates for DAPT versus aspirin alone were estimated by Kaplan-Meier methods and compared with the log-rank test. Cox proportional hazards models were created to evaluate correlates of moderate or severe bleeding; the response variable was time to first bleeding event. Separate models were created for 3 patient cohorts: overall, established vascular disease, and risk factors only. Variables associated with moderate or severe bleeding on univariable analysis (P<0.10) were entered into multivariable Cox proportional hazards models with the stepwise model selection procedure (P value for staying in the model <0.05) followed by bootstrap evaluation. A time-dependent covariate—postbaseline surgical procedures (carotid angioplasty, carotid endarterectomy, coronary artery bypass grafting, coronary angiography, coronary stent, coronary angioplasty, other coronary interventions, interventions for peripheral arterial disease, intra-aortic balloon pump, or permanent pacemaker placement) that occurred as close to the day but not after the bleeding event—was created and considered for adjustment. Multivariable interaction tests were performed to examine the relationship between age ≥75 years or weight ≤60 kg (as indicator variables) and bleeding in patients taking clopidogrel because these risk factors were significantly associated with severe bleeding in a recent trial of oral antiplatelet therapy.9 To further analyze the relationship between age and bleeding, the expected number of patients with bleeding episode, stratified by treatment, was plotted against age. The expected number is the predicted probability of bleeding at any age multiplied by the total number of patients at that age. Kaplan-Meier estimates and log-rank P values were calculated, and hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated, to analyze the relationship between moderate bleeding and all-cause mortality. Additionally, in some cases, the relationship between moderate bleeding and cardiovascular death/MI/stroke was analyzed when the time to bleeding was less than the time to end point event being analyzed for patients without fatal bleeding or intracranial hemorrhage. The effect of bleeding on mortality was assessed with the use of multivariable Cox proportional hazards models with time-dependent covariates of moderate bleeding, MI, and stroke by excluding confounding events (fatal bleeding, intracranial hemorrhage, fatal MI, or fatal stroke). The same variable selection techniques described above were used to develop the model. No adjustments for multiple comparisons were made. Statistical analyses were performed with the use of SAS software (version 9.1.3, SAS Institute).
Frequency and Severity of Bleeding
Among the 15 603 patients in CHARISMA, 487 (3.1%) suffered bleeding complications over a median of 28 months (234 [1.5%] severe; 265 [1.7%] moderately severe; 12 patients suffered both). The frequency of severe bleeding was 1.7% and 1.3% (HR, 1.25; 95% CI, 0.97 to 1.62; P=0.087) in the clopidogrel versus placebo arms; frequency of moderate bleeding was 2.1% versus 1.3% (P<0.001), respectively (Figure 1A).
The frequency of bleeding was similar in patients with established vascular disease and those with risk factors only: 1.5% and 1.6%, P=0.41 for severe bleeding; 1.7% and 1.8%, P=0.73 for moderate bleeding; and 3.1% and 3.4%, P=0.26 for moderate or severe bleeding. Because of the heterogeneity of results in CHARISMA between patients with established vascular disease (in whom there was a significant 12% relative reduction in cardiovascular death/MI/stroke) and those with risk factors only (in whom there was a nonsignificant 20% relative increase; the interaction term was marginally significant, P=0.045),6 we examined the frequency of bleeding between patients with established disease and those with risk factors only, by treatment (Figure 1B). There was no significant interaction between baseline vascular disease state and treatment for all bleeding types.
Time Course of Bleeding
In both treatment arms, the risk of moderate or severe bleeding while on clopidogrel was greatest in the first year: 2.0% versus 1.1% with placebo (HR, 1.88; 95% CI, 1.45 to 2.45; P=0.001). After the first year, among patients who had not had moderate or severe bleeding the first year, the 2 treatment groups had similar bleeding rates, and they remained stable for the remainder of the trial. The annualized frequency was 0.9% in patients on clopidogrel versus 0.8% in patients on placebo (HR, 1.18; 95% CI, 0.917 to 1.52; P=0.197). Such a trend of moderate or severe bleeding was demonstrated in the Kaplan-Meier plots in Figure 2.
Regarding recurrent bleeding, 43 patients had 2 bleeds, 4 had 3 bleeds, 1 had 4 bleeds, and 1 had 6 bleeds. Most patients had only 1 bleed (n=444). Of the 244 patients who bled during the first year of study entry, recurrent bleeding occurred in the first year in 18 patients (20 events). There were 20 recurrent bleeds in year 2 and 3 recurrent bleeds in year 3.
Study drug was discontinued in 19% (n=2999): 1415 discontinuations of placebo (9.1%), 56 because of moderate or severe bleeding, and 1584 discontinuations of clopidogrel (10.2%), 125 because of moderate or severe bleeding. During the year after enrollment, 72 patients (0.9%) stopped clopidogrel because of moderate or severe bleeding, twice as many as stopped placebo (28 patients [9.4%]; P<0.001).
Types of Bleeding
The most common source of bleeding was gastrointestinal, followed by intracranial bleeding and bleeding related to a surgical procedure (Table 1).
Baseline Correlates of Bleeding
Baseline characteristics of patients who did and did not suffer a moderate or severe bleeding complication are displayed in Table 2. Characteristics that correlated with bleeding were similar for those receiving clopidogrel and placebo, with few exceptions. In general, patients who bled were older and sicker than those who did not, regardless of treatment group.
Multivariable Analysis of Correlates of Bleeding
Multivariable analysis of correlates of bleeding risk for the overall CHARISMA population, as well as by baseline vascular disease state, is displayed in Table 3. In general, correlates of bleeding risk were similar between the overall population and the 2 subpopulations. Age was an important correlate of bleeding in both groups. Treatment with clopidogrel was also strongly associated with bleeding. However, there was no apparent interaction between age and risk of bleeding with clopidogrel when age was analyzed as either a continuous variable or dichotomously (cut point: ≥75 versus <75 years); clopidogrel did not increase bleeding more in elderly than in younger patients. There was also no apparent interaction between weight (≤60 versus >60 kg) and bleeding with clopidogrel. Heart failure was a powerful risk factor for bleeding.
Figure 3, in which the number of patients with bleeding was plotted against age, further illustrates the relationship between age and moderate or severe bleeding in patients receiving clopidogrel (Figure 3A) and placebo (Figure 3B). More bleeding occurred in younger patients. Patients treated with clopidogrel had more bleeding.
Because the vast majority of bleeding was gastrointestinal, its risk factors were assessed separately. Compared with overall bleeding (Table 3), baseline antihypertensive use and baseline statin use were no longer risk factors; the other risk factors had a similar magnitude and direction of impact on bleeding.
Frequency of Adverse Events in Patients Who Bled
We examined the frequency of primary end point events (cardiovascular death/MI/stroke) among patients who suffered a moderate bleed (Figure 4) after removing those who had fatal and intracranial bleeding. The primary end points, cardiovascular death, nonfatal MI, and all-cause death, each occurred more frequently in patients who had suffered a moderate bleed according to GUSTO criteria.
Clinical Significance of Bleeding
Bleeding and many correlates of all-cause mortality were identified on time-dependent multivariable survival analysis in patients who experienced no confounding events (fatal bleeding, intracranial hemorrhage, fatal MI, or fatal stroke). Moderate bleeding (time dependent) was powerfully and independently correlated with all-cause mortality (HR, 2.55; 95% CI, 1.71 to 3.80; P<0.0001) (Table 4). Likewise, MI (time dependent; HR, 2.92; 95% CI, 2.04 to 4.18; P<0.0001) and stroke (time dependent; HR, 5.83; 95% CI, 4.63 to 7.32; P<0.0001) were significantly associated with all-cause mortality. In a landmark analysis comparing the patients with a bleeding episode in the first 6 months with those who did not (with those who died in the first 6 months excluded), a fixed-variable proportional hazards model was constructed. The fixed covariates of moderate bleeding (HR, 2.11; 95% CI, 1.11 to 4.02), MI (HR, 1.93; 95% CI, 1.34 to 2.79), and stroke (HR, 2.68; 95% CI, 1.92 to 3.72) were still significantly associated with all-cause mortality, although not as powerfully as in the time-dependent survival model. There was no significant difference in the relationship between bleeding and death between patients with established vascular disease and those with risk factors only.
Gastrointestinal bleeding was highly associated with all-cause mortality (HR, 3.21; 95% CI, 2.21 to 4.66). In a multivariable model with the same covariate set as in Table 4, gastrointestinal bleeding retained the significance of its association with all-cause mortality (HR, 1.82; 95% CI, 1.24 to 2.69).
This analysis demonstrates that although clopidogrel significantly increased the risk of moderate or severe bleeding events, the increased the risk of severe bleeding was not significant, with an absolute incremental 0.4% risk over a median of 28 months (P=0.087). The risk of bleeding with clopidogrel was greatest during the first year, after which bleeding was no more likely to occur than in placebo-treated patients. The frequency of moderate bleeding was significantly increased by an absolute 0.8% over the same time frame (P<0.001). Moderate bleeding was strongly and independently correlated with mortality, even among patients who experienced no confounding events.
Prior studies have examined the relationship between bleeding and mortality.10 In those studies in which antiplatelet agents, anticoagulants, or fibrinolytics were administered, different definitions of bleeding were used, including not only the GUSTO definitions used in this analysis but more than a dozen others.10 A few definitions have been reported to be associated with mortality; most have not.10
Both of the most commonly used criteria to grade severity of bleeding, the GUSTO and the Thrombolysis in Myocardial Infarction (TIMI) scales, were designed to quantify clinically significant bleeding in the setting of ST-segment elevation MI treated with fibrinolytic therapy.8 Other criteria have been created for use in trials evaluating DAPT. The Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial used the Optimal Antiplatelet Strategy for Interventions (OASIS) definition; major bleeding was classified as either life threatening/ fatal, hemoglobin decrease of 5 g/dL, significant hypotension requiring inotropes or surgery, intracranial hemorrhage or necessitating transfusion of 4 U of blood; or not life threatening, requiring 2 to 3 U blood, that is significantly disabling, or intraocular, leading to loss of vision.3 In CURE, bleeding was classified using the TIMI and GUSTO criteria, allowing for comparisons with other studies. The Clopidogrel and Metoprolol in Myocardial Infarction Trial (COMMIT) used a unique definition of bleeding: life threatening, defined as hemorrhagic stroke or bleeding requiring transfusion, or fatal.2 The Clopidogrel for the Reduction of Events During Observation (CREDO) and Clopidogrel as Adjunctive Reperfusion Therapy (CLARITY) trials used TIMI criteria.1,4 CHARISMA, the fifth large placebo-controlled trial comparing aspirin and DAPT, was the only trial to analyze bleeding using GUSTO criteria and enrolling stable patients.7 In an analysis of pooled data from 2 randomized trials in which 15 454 patients with a non–ST-segment elevation acute coronary syndrome were enrolled, both GUSTO moderate or severe bleeding and TIMI major bleeding were associated with higher adjusted rates of mortality and nonfatal MI at 30 days and 6 months. After blood transfusion was included in the regression analysis, however, these outcomes were significantly associated with bleeding characterized by the GUSTO criteria but not the TIMI criteria; bleeding characterized by the GUSTO definition was associated with an increased risk of death and MI, whereas bleeding characterized by the TIMI definition was not.11 No studies of which we are aware previously identified a relationship between bleeding and mortality among stable outpatients like those enrolled in CHARISMA. We believe that this is only the second study to examine the relationship between bleeding and mortality among patients on DAPT. A prior analysis from 3 large studies (2 registries and 1 randomized clinical trial) of 34 126 patients with acute coronary syndrome on DAPT found that a major bleed defined by the CURE criteria was associated with a 5-fold mortality increase.12 The present analysis also provides a link between bleeding (defined by TIMI/GUSTO criteria) and mortality in patients receiving DAPT.
Subgroups of CHARISMA
The most puzzling result from CHARISMA was the discordant effects of clopidogrel in patients with established vascular disease versus those with risk factors only, in whom DAPT not only was not beneficial but appeared to be harmful (ie, associated with an increase in both all-cause death and cardiovascular death).7 Many wonder whether bleeding might have been more frequent in such patients, outweighing any benefit that might have been present in lower-risk patients. Patients with established vascular disease and those with risk factors were very different in many ways.6 Our results did not confirm the hypothesis that excess bleeding in lower-risk patients explained their greater mortality. Moderate or severe bleeding was not significantly greater in patients with risk factors alone than in those with established disease (3.1% versus 3.4%, respectively; P=0.260). Thus, given that the potential benefit from DAPT therapy in patients with risk factors alone is much smaller than in those with established disease, the increased risk of bleeding argues against the administration of DAPT in this population.
Patients at increased risk of bleeding were excluded from CHARISMA; a registry of excluded patients was not maintained. The risk of bleeding with aspirin and clopidogrel would undoubtedly be higher in unselected patients; the results of this study should not be generalized to patients unlike those enrolled. Despite the large size of the trial, there is limited ability to analyze correlates of and risk associated with different sites and types of bleeding because of relatively few bleeding events that were not gastrointestinal. Similarly, there are too few data to analyze the risk factors and clinical impact of recurrent bleeding. There were many differences between patients who bled and did not bleed. The ability to accurately identify independent correlates of mortality between 2 groups is reduced when there are many differences between them. We did not collect information about proton pump inhibitor or H2 blocker use; 2 recent randomized trials have recently indicated that proton pump inhibitors reduce peptic ulcers and gastrointestinal complaints in patients on aspirin, and a third indicated the same for patients on aspirin and clopidogrel, without suggestion of harm.13–15 If proton pump inhibitors do reduce gastrointestinal bleeding, they would likely only reduce upper gastrointestinal bleeding. Although doses of aspirin between 75 and 162 mg were permitted, lower doses of aspirin with this narrow range are associated with lower rates of bleeding.16 Another limitation is that multiple analyses were performed in this study, and no adjustments for multiple comparisons were made.
Clopidogrel increases bleeding in stable patients like those enrolled in CHARISMA. The excess risk is largely confined to the first year of therapy. Moderate bleeding characterized by the GUSTO scale is independently associated with all-cause and cardiovascular mortality. Excess bleeding did not explain the increase in mortality associated with clopidogrel in patients with risk factors alone. Characteristics associated with the risk of bleeding have been identified that should enable clinicians to make more informed decisions about who might benefit (or be harmed) by the prolonged administration of DAPT.
We would like to thank Yvette Henry for excellent editorial assistance.
Sources of Funding
The CHARISMA study was funded by Sanofi-Aventis and Bristol-Myers Squibb.
Dr Berger has served as a consultant for Accumetrics, Boehringer Ingelheim, Eli Lilly/Daiichi-Sankyo, Novartis/Portola, AstraZeneca, and Guerbet (each for <$10 000) and has received research funding from Helena, Haemoscope, and Corgenix/Aspirinworks (for ≤$10,000), Accumetrics, AstraZeneca, The Medicines Company, Thrombovision, and Lilly/Daiichi-Sankyo (for >$10,000). He has had full access to all of the data in the study and takes responsibility for the integrity and accuracy of the analysis. Dr Bhatt has served as the Principal Investigator of the CHARISMA trial; no personal compensation was received for this. Research grants (to the institution) were received from AstraZeneca, Bristol-Myers Squibb, Eisai, Ethicon, Heartscape, Sanofi-Aventis, The Medicines Company, and AstraZeneca; he has served as a Consultant/Advisory Board member (honoraria waived or donated) for Arena, Astellas, AstraZeneca, Bayer, Bristol-Myers Squibb, Cardax, Centocor, Cogentus, Daiichi-Sankyo, Eisai, Eli Lilly, GlaxoSmithKline, Johnson & Johnson, McNeil, Medtronic, Millennium, Molecular Insights, Otsuka, Paringenix, PDL, Philips, Portola, Sanofi-Aventis, Schering-Plough, Scios, Takeda, The Medicines Company, and Vertex. M. Shao reports no conflicts. D.M. Brennan reports no conflicts. Prof Hacke was a member of the CHARISMA Executive Committee and was reimbursed for his time and efforts. He also is a member of the Sanofi-Aventis International Advisory Board. He has no additional potential conflicts of interest to declare. Dr Montalescot reports having received research grants from Bristol-Myers Squibb, Sanofi-Aventis Group, Eli Lilly, Guerbet Medical, Medtronic, Boston Scientific, Cordis, Stago, Centocor, Fondation de France, INSERM, Fédération Française de Cardiologie, Société Française de Cardiologie. He reports having received consulting fees from Sanofi-Aventis Group, Eli Lilly, Bristol-Myers Squibb, The Medicines Company, and Schering-Plough and lecture fees from Sanofi-Aventis, Eli Lilly, Bristol-Myers Squibb, Merck Sharp & Dohme, Cordis, GlaxoSmithKline, and Schering-Plough. Dr Steinhubl is currently employed by The Medicines Company. In the prior 2 years, he has received honoraria as a consultant for AstraZeneca, Daiichi-Sankyo, Eli Lilly, Sanofi-Aventis, Bristol-Myers Squibb, Portola, Arena, Cogentus, Johnson & Johnson, and Cardax. Dr Topol served as the chair of the CHARISMA trial and did not receive personal compensation for this role. He has served as a consultant to Daiichi-Sankyo, Portola Pharmaceuticals, and Sanofi-Aventis. Dr Fuster reports no conflicts. Dr Steg reports receiving a research grant (to the institution) from Sanofi-Aventis and having received honoraria for speaking or consulting from Astellas, AstraZeneca, Bayer, Boehringer-Ingelheim, Bristol-Myers Squibb, Endotis, GlaxoSmithKline, Medtronic, Merck Sharp & Dohme, Nycomed, Sanofi-Aventis, Servier, and The Medicines Company. Dr Fox reports having received grants and honoraria from Sanofi-Aventis, GlaxoSmithKline, and Bristol-Myers Squibb and a grant from Lilly.
Sabatine MS, Cannon CP, Gibson CM, Lopez-Sendon JL, Montalescot G, Theroux P, Claeys MJ, Cools F, Hill KA, Skene AM, McCabe CH, Braunwald E; CLARITY-TIMI 28 Investigators. Addition of clopidogrel to aspirin and fibrinolytic therapy for myocardial infarction with ST-segment elevation. N Engl J Med. 2005; 352: 1179–1189.
Chen ZM, Jiang LX, Chen YP, Xie JX, Pan HC, Peto R, Collins R, Liu LS; COMMIT (ClOpidogrel and Metoprolol in Myocardial Infarction Trial) Collaborative Group. Addition of clopidogrel to aspirin in 45 852 patients with acute myocardial infarction: randomised placebo-controlled trial. Lancet. 2005; 366: 1607–1621.
Steinhubl SR, Berger PB, Mann JT III, Fry ET, DeLago A, Wilmer C, Topo EJ; CREDO Investigators (Clopidogrel for the Reduction of Events During Observation). Early and sustained dual oral antiplatelet therapy following percutaneous coronary intervention: a randomized controlled trial. JAMA. 2002; 288: 2411–2420.
Grines CL, Bonow RO, Casey DE Jr, Gardner TJ, Lockhart PB, Moliterno DJ, O'Gara P, Whitlow P. Prevention of premature discontinuation of dual antiplatelet therapy in patients with coronary artery stents: a science advisory from the American Heart Association, American College of Cardiology, Society for Cardiovascular Angiography and Interventions, American College of Surgeons, and American Dental Association, with representation from the American College of Physicians. Circulation. 2007; 115: 813–818.
Bhatt DL, Fox KA, Hacke W, Berger PB, Black HR, Boden WE, Cacoub P, Cohen EA, Creager MA, Easton JD, Flather MD, Haffner SM, Hamm CW, Hankey GJ, Johnston SC, Mak KH, Mas JL, Montalescot G, Pearson TA, Steg PG, Steinhuble SR, Weber MA, Brennan DM, Fabry-Riboudo L, Booth J, Topol EJ; CHARISMA Investigators. Clopidogrel and aspirin versus aspirin alone for the prevention of atherothrombotic events. N Engl J Med. 2006; 354: 1706–1717.
Bhatt DL, Flather MD, Hacke W, Berger PB, Black HR, Boden WE, Cacoub P, Cohen EA, Creager MA, Easton JD, Hamm CW, Hankey GJ, Johnston SC, Mak K-H, Mas J-L, Montalescot G, Pearson TA, Steg PG, Steinhubl SR, Weber MA, Fabry-Ribaudo L, Hu T, Topol EJ, Fox KAA; for the CHARISMA Investigators. Patients with prior myocardial infarction, stroke, or symptomatic peripheral arterial disease in the CHARISMA trial. J Am Coll Cardiol. 2007; 49: 1982–1988.
Wiviott SD, Braunwald E, McCabe CH, Montalescot G, Ruzyllo W, Gottlieb S, Neumann F-J, Ardissino D, DeServi S, Murphy SA, Riesmeyer J, Weerakkody G, Gibson CM, Antman EM; for the TRITON-TIMI38 Investigators. Prasugrel versus clopidogrel in patients with acute coronary syndromes. N Engl J Med. 2007; 357: 2001–2015.
Steinhubl SR, Kastrati A, Berger PB. Variation in the definitions of bleeding in clinical trials of patients with acute coronary syndromes and undergoing percutaneous coronary interventions, and its impact on the apparent safety of antithrombotic drugs. Am Heart J. 2007; 154: 3–11.
Rao SV, O'Grady K, Pieper KS, Granger CB, Newby LK, Mahaffey KW, Moliterno DJ, Lincoff AM, Armstrong PW, Van de Wert F, Califf RM, Harrington RA. A comparison of the clinical impact of bleeding measured by two different classifications among patients with acute coronary syndromes. J Am Coll Cardiol. 2006; 47: 809–816.
Eikelboom JW, Mehta SR, Anand SS, Xie C, Fox KAA, Yusuf S. Adverse impact of bleeding on prognosis in patients with acute coronary syndromes. Circulation. 2006; 114: 774–782.
Scheiman JM, van Rensburg CJ, Uebel P, Douda L, Niezgoda K, Naucler E, Svedberg L-E. Prevention of low-dose acetylsalicylic-associated gastric/duodenal ulcers with esomeprazole 20mg and 40mg once daily in patients at increased risk of ulcer development: a randomized controlled trial (OBERON). Gastroenterology. 2009; 136 (suppl 1): A-70.
Yeomans ND, Lanas A, Labenz J, van Zanten SV, van Rensberg C, Rácz I, Tchernev K, Karamanolis D, Roda E, Hawkey C, Nauclér E, Svedberg L-E. Efficacy of esomeprazole (20 mg once daily) for reducing the risk of gastroduodenal ulcers associated with continuous use of low-dose aspirin. Am J Gastroenterol. 2008; 103: 2465–2473.
Bhatt DL. COGENT: a prospective, randomized, placebo-controlled trial of omeprazole in patients receiving aspirin and clopidogrel. Presented at: Transcatheter Cardiovascular Therapeutics; September 24, 2009; San Francisco, Calif.
Steinhubl SR, Bhatt DL, Brennan DM, Montalescot G, Hankey GJ, Eikelboom JW, Berger PB, Topol EJ; on behalf of the CHARISMA Investigators. Aspirin to prevent cardiovascular disease: the association of aspirin dose and clopidogrel with thrombosis and bleeding. Ann Intern Med. 2009; 150: 379–386.
Choosing the optimal antiplatelet regimen for a patient requires a thorough understanding of not only the benefits of therapy but the risks as well. This is especially important now that several different antiplatelet agents are available. The Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trial is the largest and longest study of the risks of dual antiplatelet therapy, and this analysis of CHARISMA data provides insights into the frequency of bleeding, risk factors for bleeding, and consequences of bleeding. Clinicians will find the data useful when they try to balance the risks and benefits of dual antiplatelet therapy with aspirin and clopidogrel in stable patients with vascular disease or with risk factors for vascular disease, like those enrolled in the CHARISMA trial.
Continuing medical education (CME) credit is available for this article. Go to http://cme.ahajournals.org to take the quiz.
Guest Editor for this article was James T. Willerson, MD.