A subset of patients recruited in the main MATRIX study will be randomized after intervention but before discharge to standard of care (the treating physician will decide which oral P2Y12 inhibitor will be added on top of aspirin) versus a customized approach based on an algorithm which integrates phenotypic information, including but not limited to residual on-treatment platelet reactivity assessed via VerifyNow P2Y12 Assay.

Up to 20-30% of clopidogrel treated patients do not adequately respond to the drug and are at higher risk for ischemic events including death, myocardial infarction, stroke and stent thrombosis.

Residual high on-treatment platelet reactivity while the patient is on clopidogrel depends on a complex interplay of phenotypic (spontaneous platelet reactivity, inflammatory status, acuity of the clinical presentation, age, renal function) and genetic variables.

Two main Loss of function alleles have been identified: 1) CYP450 2C19*2 is present in around 25% of the Caucasian population and result in a lower amount of clopidogrel active metabolite. Carriers of 2C19*2 are at higher risk for death or MI and 2.7 fold increase in the risk of stent thrombosis if treated with conventional clopidogrel; 2) ABCB-1 C carriers have reduced clopidogrel absorption and they have similarly been shown to be at higher risk for ischemic adverse events if treated with clopidogrel. Many investigators have recently shown however, that the positive predictive value of genetic testing alone at the time of PCI is limited and the knowledge of genetic status alone with respect to the two previously described loss of function alleles is only poorly able to identify to long-term clopidogrel poor responders. An Algorithm has therefore been developed, combining phenotype information which has been shown to risk stratify both ischemic and bleeding events up to one year follow-up in PCI patients.

This algorithm has been developed from a single center retrospective registry. To prospectively validate it in the context of a prospective multicenter study, the first 320 patients recruited in the present study will undergo phenotype at discharge and at 30 days and genotype assessment at the time of randomization, irrespective of the group which they have been assigned to (i.e. standard of care or gene and phenotype). The hypothesis behind this mechanistic sub-study is that the use of this combined phenotype-genotype algorithm will increase the proportion of patients at 30 days who will be in the therapeutic range according to PRU values from 50% in the standard of care versus 70% in the gene and phenotype group.

Inclusion Criteria:

• patients recruited in the main MATRIX study who underwent coronary angioplasty with stent placement.

Exclusion Criteria:

• unwillingness to sign this sub study specific informed consent