Clinical outcomes according to permanent discontinuation of clopidogrel or placebo in the CHARISMA trial

Abstract

Background: Late discontinuation of clopidogrel after an acute coronary syndrome or stent placement may be associated with a clinical rebound effect.

Aims: To describe the characteristics and evolution of patients non-compliant to study drug in the prospective, randomized, double-blind CHARISMA trial.

Methods: Of 15,603 patients aged 45 or older years with established atherothrombotic disease (coronary artery disease, stroke, peripheral arterial disease) or multiple cardiovascular risk factors, 2999 permanently interrupted (withdrawers) study drug (clopidogrel or placebo) during follow-up. The primary endpoint was first occurrence since randomization of myocardial infarction, stroke or cardiovascular death.

Results: Withdrawers displayed a higher risk profile and rates of death/myocardial infarction/stroke (13.5% versus 5.6%; hazard ratio [HR]: 3.18; 95% confidence interval [CI]: 3.05-3.32; p<0.001) and severe bleeding (4.9% versus 0.7%; odds ratio [OR]: 7.42; 95% CI: 5.67-9.70; p<0.001) versus non-withdrawers. Death/myocardial infarction/stroke occurred after an average of 228 days (95% CI: 197-258) and was less frequent in patients assigned to clopidogrel versus placebo (9.7% versus 11.9%; HR: 0.80; 95% CI: 0.64-1.00; p=0.051); the rate of severe bleeding was the same (4.0% versus 4.3%; OR: 0.92; 95% CI: 0.65-1.32; p=0.66). Among withdrawers, initial clopidogrel treatment was an independent correlate of survival (HR: 0.74, 95% CI: 0.59-0.93; p=0.011), but not severe bleeding (OR: 0.94; 95% CI: 0.65-1.35; p=0.74). Kaplan-Meier curves for the primary endpoint suggested no rebound effect or disease reactivation after discontinuation of clopidogrel compared with placebo.

Conclusions: Patients who stopped medication had increased rates of ischaemic and bleeding events and mortality. Patients initially on clopidogrel had fewer ischaemic events than those on placebo; discontinuation was not associated with any clinically detectable rebound effect.