Bleeding with Dual-antiplatelet Therapy

Abstract & Commentary

By Michael H. Crawford, MD

Source: Berger PB, et al. Bleeding complications with dual-antiplatelet therapy among patients with stable vascular disease or risk factors for vascular disease. Circulation. 2010;121:2575-2583.

Dual-antiplatelet therapy with aspirin and clopidogrel is recommended for patients with known vascular disease or multiple risk factors for vascular disease. However, little is known about the long-term bleeding risk of such therapy. Thus, Berger and colleagues analyzed the Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance (CHARISMA) trial to determine the frequency and time course of bleeding in dual-antiplatelet-therapy patients both with and without known vascular disease. The 15,603 patients in CHARISMA were randomized to clopidogrel 75 mg/day vs. placebo; all patients took aspirin (75 to 162 mg/day). Median follow-up was 28 months. Severe bleeding (GUSTO criteria) was observed in 1.7% of the clopidogrel group vs. 1.3 % of the placebo group (p = NS). Moderate bleeding rates were 2.1% vs. 1.3%, p < 0.001. Bleeding risk was greatest in the first year on therapy. The frequency of bleeding was not related to the presence or absence of known vascular disease. Moderate bleeding was predictive of all-cause mortality (hazard ratio 2.55; 95% CI 1.71-3.80, p < 0.001); myocardial infarction (HR 2.92, 2.04-4.18, p < 0.0001), and stroke (HR 4.20, 3.05-5.77, p < 0.0001). The authors concluded that there is an increased risk of bleeding with long-term dual-antiplatelet therapy, which is greatest in the first year; bleeding is associated with increased mortality.


This study is of interest for three reasons: 1) it describes who should get dual-antiplatelet therapy based upon a risk vs. benefit analysis; 2) it quantifies the risk of bleeding; and 3) it establishes that moderate bleeding is associated with increased mortality. The overall CHARISMA study results showed that the benefits of dual-antiplatelet therapy were heterogeneous. In patients with established vascular disease, there was a significant decrease in the combined primary endpoint of death, myocardial infarction (MI), or stroke of 12%. In those with risk factors for vascular disease only, there was a 20% increase in the endpoint and the difference between the two groups was significant (p = 0.045). The risk of moderate or severe bleeding on dual therapy was double that of low-dose aspirin alone in the first year (2.0% vs. 1.1%) in both groups. Thus, the difference in major events between the two groups could not be explained by bleeding alone. After the first year, in the patients who had not bled, the bleeding rate was similar for both groups (0.8% and 0.9%). Also, 90% of patients had only one bleed. Moderate bleeding, but not severe bleeding, was associated with increased mortality. This suggests that severe bleeding is due to major pathologic processes upon which antiplatelet therapy has little influence. Moderate bleeding, on the other hand, probably represents minor pathologic processes where dual therapy can markedly augment bleeding and escalate the risk to the patient.

One important unique feature of this study is the inclusion of lower-risk patients with risk factors for vascular disease only. The results discussed above would suggest that the risk of dual therapy outweighs the potential benefit in such patients. This observation is more important when you consider that patients at increased risk for bleeding were excluded from CHARISMA. In an all-comers population, the risk of bleeding and potential harm would be expected to be higher. In CHARISMA, 19% of dual-therapy patients stopped clopidogrel due to moderate-to-severe bleeding vs. 9% in the placebo arm. Bleeding risk correlated highly with age > 75 years, heart failure, and diabetes. Most of these bleeds were gastrointestinal, so there was insufficient data to analyze the risk of other types of bleeding separately. Unfortunately, data on proton pump inhibitors (PPI) or hydrogen blockers were not obtained. Whether such agents could mitigate the risk of bleeding is unknown, but has been suggested in other studies. On the other hand, certain PPIs have been implicated in reducing the effectiveness of clopidogrel.

In most acute coronary syndrome or percutaneous coronary intervention patients without contraindications to dual-antiplatelet therapy or an increased risk of bleeding, the potential benefits of this therapy will outweigh the risks. However, in stable patients without overt vascular disease, the elderly, and those with heart failure and diabetes, the risk may outweigh the benefits. Such patients require careful clinical judgment and close follow-up if dual-antiplatelet therapy is administered. This is especially the case for prasugrel, which has a higher incidence of bleeding than clopidogrel. We are still seeking the ideal antiplatelet agent that prevents exuberant platelet activation for pathologic vascular events, yet allows appropriate platelet activation for hemostasis.