Clinical Implications of Very Low On-Treatment Platelet Reactivity in Patients Treated With Thienopyridine: The POBA Study (Predictor of Bleedings With Antiplatelet Drugs)

Abstract

This study was designed to define the hyperresponse to thienopyridine (very low on-treatment platelet reactivity [VLTPR]) as the most predictive threshold value of platelet reactivity index vasodilator-stimulated phosphoprotein (PRI VASP) for the prediction of non-access site-related bleeding events. We also aimed to identify predictors of bleeding and VLTPR in patients treated with thienopyridines. New P2Y12 blockers and platelet monitoring has been proposed to optimize platelet inhibition after acute coronary syndromes and improve ischemic outcomes. However, bleeding complications remain the Achilles' heel of antiplatelet therapy, and platelet monitoring could be useful to evaluate this risk. A total of 1,542 consecutive patients undergoing coronary stenting for ACS were included in the present study (287 taking clopidogrel 75 mg, 868 taking clopidogrel 150 mg, and 387 taking prasugrel 10 mg). During 6-month follow-up, 9% of patients (n= 139) experienced nonaccess site-related Bleeding Academic Research Consortium bleeding complications. These patients were more often women and nondiabetic and had lower PRI VASP values than others (p < 0.001). Receiver-operating characteristic curve analysis (0.71, p < 0.01) identified a threshold value for VLTPR of PRI VASP ≤10% to predict bleeding events with a sensitivity of 17% and a specificity of 97%. Although prasugrel was the main predictor of VLTPR in the whole population (odds ratio: 10.2, 95% confidence interval: 3.0 to -34.2; p < 0.001), VLTPR was the strongest and independent predictor of bleeding (odds ratio: 4.7, 95% confidence interval: 2.7 to 8.3; p < 0.001). The present study identified VLTPR (PRI VASP ≤10%) as the strongest predictor of bleeding complications in patients treated with thienopyridines. This marker could be useful for tailored therapy and bleeding prevention.