In the EARLY ACS trial, routine early eptifibatide was not superior to delayed provisional use at percutaneous coronary intervention (PCI); however, among PCI-treated patients, numerically fewer ischemic end points occurred in the upstream eptifibatide group. We sought to further explore this finding using methods for examination of treatment effect in this postrandomization subgroup.

Of 9,406 patients in the EARLY ACS primary analysis cohort, 9,166 (97.4%) underwent coronary angiography. We used Cox proportional hazards regression modeling, with PCI as a time-dependent covariate, to examine the effect of routine early versus delayed provisional eptifibatide among 5,541 patients undergoing PCI and to explore the interaction between treatment with PCI and randomized treatment strategy. After multivariable adjustment, compared with delayed provisional use, routine early eptifibatide was associated with lower rate of 30-day death or myocardial infarction (MI) after PCI (hazard ratio [HR] 0.80, 95% CI 0.68-0.95) but not with medical management (HR 0.97, 95% CI 0.74-1.29); PCI × randomized treatment interaction term P = .24. Excluding PCI-related MI, the adjusted HR for 30-day death or MI for routine early eptifibatide versus delayed provisional use was 0.80 (95% CI 0.60-1.08) for post-PCI treatment and 1.01 (95% CI 0.79-1.34) for medical management; PCI × randomized treatment interaction term P = .28.

Consistent with previous literature, upstream treatment with eptifibatide was associated with improved outcomes in high-risk non–ST-segment elevation acute coronary syndrome patients treated with PCI; however, a nonsignificant interaction term precludes a definite conclusion.