Background: While previous reports showed on-treatment ADP-induced platelet reactivity to be an independent predictor of bleeding after percutaneous coronary interventions (PCI) in stable patients, this has never been investigated in cardiogenic shock patients.

Methods: The association of bleeding events in respect to ADP-induced platelet aggregation was investigated in patients undergoing primary PCI for acute myocardial infarction complicated by cardiogenic shock (AMI-CS) and available on-treatment ADP-induced platelet aggregation measurements.

Results: In 233 patients, ADP-induced platelet aggregation was significantly lower in patients with severe bleeding classified ≥BARC3 (10 AU [IQR 3 - 13] vs. 15 AU [IQR 9 - 25], p<0.001). Multivariate analysis identified on-treatment ADP-induced platelet aggregation as an independent risk factor for bleeding (HR=0.968 per AU, 95% confidence interval (CI) 0.942-0.994). An optimal cutoff value of <12 AU for ADP-induced platelet aggregation to predict BARC ≥3 bleeding was identified via ROC analysis. Moreover, use of VA-ECMO (HR 1.972, 95% CI 1.003-3.879) or coaxial left ventricular assist device (HR 2.593, 95% CI 1.509-4.455), first lactate on ICU (HR 1.093 per mmol/l, 95% CI 1.037-1.152) and thrombocyte count (HR 0.994 per G/l, 95% CI 0.990-0.998) were independent predictors of BARC ≥3 bleedings. There was no difference in survival nor in ischemic events between patients with low and high on-treatment platelet reactivity.