New risk score could predict risk of bleeding after PCI


According to a study published in Circulation Journal, a simple risk score could be used to predict which patients are at high risk of bleeding after a percutaneous coronary intervention (PCI) procedure. Additionally, a patient’s risk of bleeding (low, intermediate, or high) could be used to direct choice of treatment (eg. bare metal stent or drug-eluting stent).

Igor Mrdovic, associate professor of Medicine and Cardiology, University of Belgrade School of Medicine, Clinical Center of Serbia, Cardiology Clinic and Emergency Hospital, Belgrade, Serbia, and others wrote that bleeding following PCI in patients with acute coronary syndromes is associated with an increased of death, myocardial infarction, and stroke at 30 days, six months, and beyond. They commented: “It is noteworthy that not only major bleeding but also moderate bleeding were significantly associated with one-year mortality.”

Mrdovic et al added that an accurate bleeding prediction model “might help to determine treatment strategies in patients presenting with ST-segment elevation myocardial infarction (STEMI)” as measures aimed at reducing the risk of bleeding might led to a reduction in an adverse outcomes. However, they explained, at present, there is not a simple risk score for predicting the risk of bleeding in STEMI patients undergoing PCI. They wrote: “We therefore performed the present study in order to generate and validate a simple risk model for the prediction of bleeding after PCI.”

Using data from the RISK-PCI, which was a single-centre study designed to predict the risk of major adverse cardiac events at 30 days in patients treated with clopidogrel, Mrdovic et al identified five factors that were associated with an increased risk of 30-day bleeding: female gender, history of peptic ulcer, creatinine clearance (<60ml/min) at time of admission to hospital, haemoglobin level (<125g/dl) at presentation, and Killip class >1 heart failure at time of admission. They then assigned points to each factor based on the risk of bleeding it conferred (eg. one point for female gender but two points for creatinine clearance).

A patient’s risk of bleeding was then calculated from their total score of points (ranging from zero to eight), with zero points indicating a low risk of bleeding, one to two points indicating a intermediate risk of bleeding, and ≥3 points indicating a high risk of bleeding. The authors reported that there was a graded 33-fold increase in 30-day bleeding with increasing risk score (p<0.001). They added that the score was predictive of both access site bleeding and bleeds not associated with the access site and commented: “An 11-fold graded increase in the primary endpoint [type ≥3 bleed in accordance with Bleeding Academic Research definition not associated with coronary artery bypass grafting] was observed between patients in the low-risk class and those in the high-risk class (p[trend]<0.001).” The authors validated their findings with data from the ART-PCI trial, which investigated the impact of high on-treatment platelet reactivity on clinical outcomes in PCI patients.

Mrdovic et al wrote that the implications of their risk model were that it could be used direct treatment. They stated: “Patients in the low-risk class might be treated via femoral approach, using a bare metal stent or drug-eluting stent and optimal doses of antiplatelet drugs. In contrast, patients in the high-risk class should be treated preferably with bare metal stents and/or a radial approach, which appeared to be associated with a significant reduction in vascular complications compared with the femoral approach.” They added that, in high-risk patients, excessive does of antithrombotic drugs should be avoided as should the potent thienopyriddines.