Phenotyping vs. genotyping for prediction of clopidogrel efficacy and safety: the PEGASUS-PCI study

Abstract

Background: Prognostic values of genotyping and phenotyping for assessment of clopidogrel responsiveness have been shown in independent studies. Objectives: To compare different assays for prediction of events during long-term follow-up. Methods: In this prospective cohort study polymorphisms of CYP2C19*2 and CYP2C19*17 alleles, vasodilator- stimulated phosphoprotein phosphorylation (VASP) assay, multiple electrode aggregometry (MEA), cone and platelet analyser (CPA) and platelet function analyser (PFA- 100) were performed in 416 patients undergoing percutaneous coronary intervention. The rates of events were recorded during a 12-month follow-up. Results: Platelet aggregation by MEA predicted stent thrombosis (2.4%) better (c-index = 0.90; P < 0.001; sensitivity = 90%; specificity = 83%) than the VASP assay, CPA or PFA-100 (c-index < 0.70; P > 0.05; sensitivity < 70%; specificity < 70% for all) or even the CYP2C19*2 polymorphism (c-index < 0.56; P > 0.05; sensitivity = 30%; specificity = 71%). Survival analysis indicated that patients classified as poor responders by MEA had a substantially higher risk of developing stent thrombosis or MACE than clopidogrel responders (12.5% vs. 0.3%, P < 0.001, and 18.5% vs. 11.3%, P = 0.022, respectively), whereas poor metabolizers (CYP2C19*1/*2 or *2/*2 carriers) were not at increased risks (stent thrombosis, 2.7% vs. 2.5%, P > 0.05; MACE, 13.5% vs. 12.1%, P = 0.556). The incidence of major bleedings (2.6%) was numerically higher in patients with an enhanced vs. poor response to clopidogrel assessed by MEA (4% vs. 0%) or in ultra-metabolizers vs. regular metabolizers (CYP2C19*17/*17 vs. CYP2C19*1/*1; 9.5% vs. 2%). The classification tree analysis demonstrated that acute coronary syndrome at hospitalization and diabetes mellitus were the best discriminators for clopidogrel responder status. Conclusions: Phenotyping of platelet response to clopidogrel was a better predictor of stent thrombosis than genotyping.