FIELD OF THE INVENTION

The present invention relates to the field of medicine, in particular to the cardiovascular field.

BACKGROUND OF THE INVENTION

Percutaneous coronary intervention (PCI) with stent implantation has become the standard of care for myocardial revascularization, especially in the setting of unstable coronary artery disease (CAD). Despite the use of dual antiplatelet therapy (DAPT; aspirin and clopidogrel), which reduces cardiovascular events after PCI by more than 80%, definite stent thrombosis (ST) remains a concern. Although it is a rare complication occurring in 0.5- 4% of patients within the first year of percutaneous coronary intervention, the vast majority of stent thrombosis cases take place in the first month and are defined as early stent thrombosis (EST). It is a devastating complication with a mortality rate up to 40% and large myocardial infarction (MI) in roughly 80% of survivors who remain exposed to frequent recurrence. Clinical and angiographic correlates of ST have been well described, of which premature interruption of DAPT is the most important risk factor for early ST. In addition, comorbidities, the initial clinical presentation, diabetes, stent undersizing or underexpansion, complex and/or bifurcation lesions, and coronary dissection have been associated with ST.

Much attention has recently been focused on patient response to clopidogrel with a strong relationship between high on-clopidogrel platelet reactivity and stent thrombosis despite stringent adherence to DAPT (Bonello et al., 2010a). It has repeatedly been shown that variable or insufficient clopidogrel active metabolite generation is the primary explanation for poor responsiveness to clopidogrel (Bonello et al, 2010a). There is a strong link between CYP2C19 genetic polymorphisms, decreased clopidogrel responsiveness measured by platelet function assays, and adverse clinical outcomes (Collet et al, 2009; Hulot et al, 2010; Mega et al, 2010a; Simon et al, 2009). Other genes might also influence clopidogrel absorption and metabolism (Bouman et al. 2010; Kazui et al., 2009; Mega et al, 2010b), but their relative contribution to the occurrence of stent thrombosis has been rarely explored. Additionally, a number of variants in genes encoding key factors of the coagulation/fibrino lytic systems and platelet receptor function have been associated with a higher risk of arterial thrombotic events but their relation to ST remains unknown (Klerk et al, 2002; Satra et al, 2011; Wang et al, 2006).

The population needing an appropriate antiplatelet therapy after PCI is about 150,000 patients per year in France and more than 4 millions in the world. The medical issue is crucial because of the devastating complication detailed above.

Several antiplatelet therapies including a thienopyridine are now available on the market, in particular clopidogrel (Plavix®), prasugrel (Efient®) and ticagrelor (Brilique®). Some of these therapies reduce the mortality but present increased risk of hemorrhages. Generic drugs will be soon on the market.

Therefore, there is a strong need of tools for helping the physician to choose the most appropriate treatment for each patient, based on clinical and economic considerations.

SUMMARY OF THE INVENTION

The inventors firstly identified three genes (CYP2C19, ABCB1, ITGB3) allowing the determination, for a given subject with a clopidogrel treatment, of being at risk of developing early stent thrombosis. Indeed, four genetic variants in three genes were found to be independently associated with the occurrence of early stent thrombosis. In addition to the genetic test, the determination of the risk may be more accurately assessed by combining the genetic test with some particular clinical risk factors. Based on the EST risk determination, the physician may choose the most appropriate treatment for each patient.

Therefore, the present invention concerns a method for determining if a subject is at risk of developing early stent thrombosis with a clopidogrel treatment, which comprises: a) genotyping in a sample from the subject the single nucleotide polymorphisms (SNP) rs4244285 and rsl2248560 of CYP2C19 gene, rsl045642 of ABCB1 and rs5918 ofiTG5J;

b) combining said SNPs genotypes through a logistic function, and

c) analyzing the score of said logistic function in order to determine the risk of developing early stent thrombosis in said subject.

Preferably, a CYP2C19 metabolic status is determined based on the genotypes of rs4244285 and rsl2248560 of CYP2C19 with the following table CYP2C19 metabolic status rs4244285 rsl2248560

Rapid GG CT and TT

Normal AG CT

GG CC

Intermediate AG CC

Slow AA CC

and the lowest is the CYP2C19 metabolic status, the highest is the risk of developing early stent thrombosis.

Preferably, the genotype TT of the rs 1045642 of ABCB1 and the genotypes TT and TC of rs5918 of ITGB3 are associated with a higher risk of developing early stent thrombosis.

Preferably, the weight factor of each gene among CYP2C19, ABCB1 and ITGB3 is about the same.

In a very specific embodiment, the logistic function is the following

(0.7196 * CYP2C19 metabolic status) + (0.7191 * ABCB1 rsl045642 genotype) - (0.7277 * ITGB3 rs5918 genotype) - 1.5838

wherein the value of CYP2C19 metabolic status is 0 when rapid, 1 when normal, 2 when intermediate, 3 when slow;

wherein the value ABCB1 rs 1045642 genotype is 1 when the genotype is TT and 0 when the genotype is CC or CT;

wherein the value ITGB3 rs5918 genotype is 1 when the genotype is CC or CT and 0 when the genotype is TT.

With this logistic function, a score of less than -0.8723 is indicative of a low risk of early stent thrombosis for the subject, a score between -0.8723 to -0.1527 is indicative of an intermediate risk of early stent thrombosis for the subject, and a score of more than -0.1527 is indicative of a high risk of early stent thrombosis for the subject.

Preferably, the method further comprises the determination of at least one clinical parameter selected from the group consisting of Type C lesion, proton pump inhibitor, diabetes mellitus, left ventricular dysfunction, percutaneous coronary intervention (PCI) in acute setting and clopidogrel loading dose, preferably the determination of all of these clinical parameters; the combination of said genotypes of the SNPs and said clinical parameters through a weighted logistic function, and analyzing the score of said logistic function in order to determine the risk of developing early stent thrombosis in said subject.

In another very specific embodiment, the logistic function is the following (0.7825 * proton pump inhibitor) + (0.6991 * CYP2C19 metabolic status) - (0.6656 * ITGB3 rs5918 genotype) + (0.6133 * diabetes) + (0.8348 * left ventricular dysfunction) + (0.7767 * ABCBl rsl045642 genotype) + (0.8459 * type C lesion) + (1.0896 * acute setting) - (0.3086 * clopidogrel loading dose) - 2.9262

wherein the value of CYP2C19 metabolic status is 0 when rapid, 1 when normal, 2 when intermediate, 3 when slow;

wherein the value ABCBl rs 1045642 genotype is 1 when the genotype is TT and 0 when the genotype is CC or CT;

wherein the value ITGB3 rs5918 genotype is 1 when the genotype is CC or CT and 0 when the genotype is TT.

wherein the value 1 is given for the presence of each of the following parameters Type C lesion, proton pump inhibitor, diabetes mellitus, left ventricular dysfunction, percutaneous coronary intervention (PCI) in acute setting and the 0 when absent; and

wherein a value of 0 is given for a clopidogrel loading dose of less than 150mg, a value of 1 when the dose is between 150 and 300 mg, and a value of 2 when the dose is between 300 and 600mg, and a value of 3 when the dose is higher than 600 mg.

With this logistic function, a score of less than -1.4126 is indicative of a low risk of early stent thrombosis for the subject, a score between -1.4126 to -0.4349 is indicative of an intermediate risk of early stent thrombosis for the subject, and a score of more than -0.4349 is indicative of a high risk of early stent thrombosis for the subject.

The risk of early stent thrombosis for the subject with a clopidogrel treatment is useful for selecting the most appropriate treatment after percutaneous coronary intervention (PCI) with stent implantation. Preferably, the treatment is to be selected among clopidogrel, prasugrel, and ticagrelor.

The present invention also concerns a kit for determining if a subject with a clopidogrel treatment is at risk of developing early stent thrombosis comprising means for genotyping the single nucleotide polymorphisms (SNP) rs4244285 and rsl2248560 of CYP2C19 gene, rsl045642 of ABCBl and rs5918 of ITGB3. Preferably, the kit further comprises instructions allowing the determination of the risk of early stent thrombosis for the subject.

The present invention further concerns the use of a kit for genotyping the single nucleotide polymorphisms (SNP) rs4244285 and rsl2248560 of CYP2C19 gene, rsl045642 of ABCBl and rs5918 of ITGB3 for determining the risk of developing early stent thrombosis in a subject with a clopidogrel treatment. BRIEF DESCRIPTION OF THE DRA WINGS

Figure 1. Study Flow Diagram

Figure 2. Distribution of CYP2C19 metabolism status between early ST and controls as predicted by genotype.

Figure 3. Proportion of patients with early ST according to the risk allelic count score.

The score was calculated by adding the number of alleles at risk (Fig 3A). The total score ranges from 1 (lowest number of alleles at risk) to 8 (highest number of alleles at risk) points. The total number of patients is reported for each group (Fig 3B).

Figure 4. Independent predictors of early stent thrombosis.

Figure 5. Area Under the Receiver Operating Characteristic Curve for Prediction of early ST. Clinical model based on nongenetic predictors (Type C lesion, proton pump inhibitor, diabetes mellitus, left ventricular dysfunction <35%, PCI in acute setting, clopidogrel loading dose) sensitivity: 60%, specificity 70%, likelihood ratio: 2.1; Genetic model containing CYP2C19 metabolic status, ABCB1 3435 TT genotype and ITGB3 PLA 2 polymorphism, sensitivity=48%>, specificity=78%>, likelihood ratio=2.0; Complete model containing all clinical, angiographic and genetic predictors. Sensitivity=67%, Specificity=79% Likelihood ratio=3.4.

Figure 6. Estimation by Odds ratios of the risk of early stent thrombosis of the population according to each model (genetic, clinical or comprehensive) and presented in tertile.

DETAILED DESCRIPTION OF THE INVENTION

The present invention provides tools for determining, in a given subject, the risk of an early stent thrombosis with a treatment by clopidogrel. Based on the risk, a therapeutic decision may be taken by the physician in order to select the most appropriate treatment for each individual.

Definition

By "subject", it is herein preferably referred to a human, more particular a human necessitating a stent implantation.

By "early stent thrombosis" is intended stent thrombosis occurring within a period of 30 days after the surgical intervention, more particular the stent implantation.

"stent thrombosis" is well known to one of skill in the art and is characterized by occlusion of the light of the stent by thrombotic platelets material. By "subject at risk of developing early stent thrombosis" means that the subject is predisposed to develop early stent thrombosis or that the subject presents an increased risk of developing early stent thrombosis with clopidogrel treatment with respect to a general population of individuals.

The present invention arises from the identification, by the inventors, of four genetic variants of three genes, namely CYP2C19, ABCBl, and ITGB3, which are strongly and independently associated with the occurrence of the early stent thrombosis (EST) with a clopidogrel treatment. These four genetic variants are:

- CYP2C19*2: rs4244285 of CYP2C19;

- CYP2C19*17: rsl2248560 of CYP2C19;

- ABCBl TT: rsl045642 of ABCBl; and

- ITGB3 PLA2/PLA2: rs5918 ΐ!ΤβΒ3.

By "CYP2C19 rs4244285 genotype" is intended the genotype of the SNP rs4244285 located in the CYP2C19 gene and defined in SEQ ID No 1. By "CYP2C19*2" is intended the allele A of the SNP rs4244285.

By "CYP2C19 rsl2248560 genotype" is intended the genotype of the SNP rsl2248560 located in the ABCBl gene and defined in SEQ ID No 2. By "CYP2C19* 17" is intended the allele T of the SNP rsl2248560.

By "ABCBl rsl045642 genotype" is intended the genotype of the SNP rsl045642 located in the ABCBl gene and defined in SEQ ID No 3.

By "ITGB3 rs5918 genotype" is intended the genotype of the SNP rs5918 located in the ITGB3 gene and defined in SEQ ID No 4.

Accordingly, the present invention concerns a method for determining if a subject with a clopidogrel treatment is at risk of developing early stent thrombosis, which comprises:

a) genotyping in a sample from the subject the single nucleotide polymorphisms (SNP) rs4244285 and rsl2248560 of CYP2C19 gene, rsl045642 of ABCBl and rs5918 ofiTG5J; b) combining said SNPs genotypes through a logistic function, and c) analyzing the score of said logistic function in order to determine the risk of developing early stent thrombosis in said subject. Sample and Genotyping

The sample may be any biological sample derived from a subject, which contains nucleic acids. Examples of such samples include fluids, tissues, cell samples, organs, biopsies, and the like. Most preferred samples are blood, plasma, saliva, jugal cells, urine, seminal fluid. In a particularly preferred embodiment, the sample is blood, plasma or saliva, more preferably saliva.

The method may include a previous step of providing a sample from the subject. The sample may be collected according to conventional techniques and used directly for diagnosis or stored.

The sample may be treated prior to performing the method, in order to render or improve availability of nucleic acids or polypeptides for testing. Treatments include, for instance, lysis (e.g., mechanical, physical, chemical, etc.), centrifugation, etc. Also, the nucleic acids may be pre-purified or enriched by conventional techniques, and/or reduced in complexity. Nucleic acids may also be treated with enzymes or other chemical or physical treatments to produce fragments thereof. Considering the high sensitivity of the claimed methods, very few amounts of sample are sufficient to perform the assay.

The sample is preferably contacted with reagents such as probes, or primers in order to assess the genotype of the SNPs. Contacting may be performed in any suitable device, such as a plate, tube, well, glass, etc. In specific embodiments, the contacting is performed on a substrate coated with the reagent, such as a nucleic acid array. The substrate may be a solid or semi-solid substrate such as any support comprising glass, plastic, nylon, paper, metal, polymers and the like. The substrate may be of various forms and sizes, such as a slide, a membrane, a bead, a column, a gel, etc. The contacting may be made under any condition suitable for a complex to be formed between the reagent and the nucleic acids of the sample.

The genotype of the SNPs may be detected by sequencing, selective hybridisation and/or selective amplification.

Sequencing can be carried out using techniques well known in the art, using automatic sequencers. The sequencing may be performed on the complete genes or, more preferably, on specific domains thereof, typically those known or suspected to carry deleterious mutations or other alterations. Amplification is based on the formation of specific hybrids between complementary nucleic acid sequences that serve to initiate nucleic acid reproduction.

Amplification may be performed according to various techniques known in the art, such as by polymerase chain reaction (PCR), ligase chain reaction (LCR), strand displacement amplification (SDA) and nucleic acid sequence based amplification (NASBA). These techniques can be performed using commercially available reagents and protocols. Preferred techniques use allele-specific PCR or PCR-SSCP. Amplification usually requires the use of specific nucleic acid primers, to initiate the reaction.

Hybridization detection methods are based on the formation of specific hybrids between complementary nucleic acid sequences that serve to detect nucleic acid sequence alteration(s).

A particular detection technique involves the use of a nucleic acid probe specific for each allele of the gene, followed by the detection of the presence of a hybrid. The probe may be in suspension or immobilized on a substrate or support (as in nucleic acid array or chips technologies). The probe is typically labelled to facilitate detection of hybrids.

In a most preferred embodiment, a gene variation is determined by DNA chip analysis. Such DNA chip or nucleic acid microarray consists of different nucleic acid probes that are chemically attached to a substrate, which can be a microchip, a glass slide or a microsphere- sized bead. A microchip may be constituted of polymers, plastics, resins, polysaccharides, silica or silica-based materials, carbon, metals, inorganic glasses, or nitrocellulose. Probes comprise nucleic acids such as cDNAs or oligonucleotides that may be about 10 to about 60 base pairs. To determine the alteration of the genes, a sample from a test subject is labelled and contacted with the microarray in hybridization conditions, leading to the formation of complexes between target nucleic acids that are complementary to probe sequences attached to the microarray surface. The presence of labelled hybridized complexes is then detected. Many variants of the microarray hybridization technology are available to the man skilled in the art (see e.g. the review by Kidgell&Winzeler, 2005 or the review by Hoheisel, 2006).

More preferably, a method allowing a rapid genotyping of the SNPs will be carried out, preferably a method allowing the genotyping in less than 1 or 2 hours. For instance, genetic testing devices are already commercially available for CYP2C19*2 (rs4244285) as Spartan ® of Spartan Bioscience Inc or for CYP2C19 as Verigene® of Nanosphere Inc. Genetic test

The logistic function is obtained by combining the relative weight of each predictor, as individually determined in the logistic regression, to an algorithm value associated to the absence or presence of the factor (binary variables) or to a classification (categorical varialbes). A negative sign is used when the parameters harbor a negative correlation with the occurrence of early stent thrombosis and a positive sign when there is a positive correlation. The quality of the logistic function may be analyzed with the aid of a ROC curve. The method for determining the logistic function is well-known by the one of skill in the art. The score is then calculated by summing the weighted predictors in the logistic function. The resulting score better reflects the combination of predictive factors and is therefore indicative of the risk, for a given subject, to develop an early stent thrombosis with a clopidogrel treatment.

Based on the SNPs genotyping of rs4244285 and rsl2248560 of CYP2C19, a CYP2C19 metabolic status may be deduced in order to recapitulate the genotyping data. In a preferred embodiment, the CYP2C19 metabolic status is determined based on the following table and a value has been assigned to each metabolic status.

Similarly, in order to prepare the logistic function, a value has been assigned to each genotype of ABCB1 and ITGB3.

Based on the genotyping, the inventors propose a logistic function such as

[al * CYP2C19 metabolic status + a2 * ABCB1 rsl045642 genotype - a3 * ITGB3 rs5918 genotype + a4] * a5

with

al, a2 and a3 being positive numbers, different to zero, selected such as their ratio a2/al, and a3/al are 1 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02; and

a4 being a number including zero; a5 is a number different to zero.

Surprisingly, the inventors observed that the statistical weights of the genes CYP2C19, ABCBl and ITGB3 are the same (see Figure 4). Indeed, the odds ratio (OR) for the CYP2C19 metabolic status, ABCBl TT and ITGB3 CC are respectively 2.01, 2.17 and 0.51. These OR of Figure 4, indicative of the statistical weights, can be used for defining the value of al, a2, a3, a4 and a5, in particular of al, a2 and a3.

In a very particular embodiment, the logistic function is the following

(0.7196 * CYP2C19 metabolic status) + (0.7191 * ABCBl rsl045642 genotype) - (0.7277 * ITGB3 rs5918 genotype) - 1.5838.

In particular, this logistic function allows the classification of the subjects in three categories of risks of early stent thrombosis with a clopidgrel treatment: a low risk (one third of the population with the lowest risk), an intermediate risk (another third of the population with the middle risk) and a high risk (one third of the population with the highest risk).

Therefore, with this logistic function, a score of less than -0.8723 is indicative of a low risk of early stent thrombosis for the subject, a score between -0.8723 to -0.1527 is indicative of an intermediate risk of early stent thrombosis for the subject, and a score of more than - 0.1527 is indicative of a high risk of early stent thrombosis for the subject.

If al, a2, a3 and a4 are different from those of the logistic function detailed above, the scores have to be adapted. For instance, if a4 is zero, then the score has to be increased by 1.5838. If the logistic function includes a5 being 2, then the score has to be multiplied by 2. Such amendments are obvious for the one skilled in the art and are included in the present invention.

Based on the information and other clinical data of the patient, the physician will decide the most appropriate treatment with an antiplatelet therapy including a thienopyridine available on the market, in particular clopidogrel (Plavix®), prasugrel (Efient®) or ticagrelor (Brilique®). For instance, the clopidogrel treatment may be avoided in the high risk population and the treatment by prasugrel or ticagrelor might be preferred. At the opposite, the clopidogrel treatment may be preferred over the treatment by prasugrel or ticagrelor in the low risk population.

Clinico-genetic test

In addition to the above detailed genetic test, the inventors defined a clinic-genetic test with higher accuracy than the genetic test alone. Indeed, they identified clinical risk factors which can be combined with the genotyping of the four SNPs in a logistic function. Accordingly, the present invention relates to a method further comprising the determination of at least one, two, three, four, five or six clinical parameters selected from the group consisting of Type C lesion, proton pump inhibitor, diabetes mellitus, left ventricular dysfunction, percutaneous coronary intervention (PCI) in acute setting and clopidogrel loading dose; the combination of said genotypes of the SNPs and said clinical parameters through a logistic function, and analyzing the score of said logistic function in order to determine the risk of developing early stent thrombosis in said subject. Preferably, the method comprises the determination of all of these clinical parameters. The one skilled in the art knows how to determine the above mentioned clinical parameters.

Therefore, the present invention concerns a method for determining if a subject with a clopidogrel treatment is at risk of developing early stent thrombosis, which comprises:

a) genotyping in a sample from the subject the single nucleotide polymorphisms (SNP) rs4244285 and rsl2248560 of CYP2C19 gene, rsl045642 of ABCB1 and rs5918 ofiTG5J;

b) determining at least one, two, three, four, five or six clinical parameters selected from the group consisting of Type C lesion, proton pump inhibitor, diabetes mellitus, left ventricular dysfunction, percutaneous coronary intervention (PCI) in acute setting and clopidogrel loading dose;

c) combining said SNPs genotypes and said clinical parameters through a logistic function, and

d) analyzing the score of said logistic function in order to determine the risk of developing early stent thrombosis in said subject.

Of course, the order of steps a) and b) may be changed.

Preferably, the method comprises the determination of all of these clinical parameters. In order to prepare the logistic function, a value may be assigned to each clinical parameter.

Clinical parameter If present If absent

AHA defined Type C coronary lesion ¹ 0

proton pump inhibitor ¹ 0

diabetes mellitus ¹ 0

left ventricular dysfunction (left ventricular ejection fration =<35%) ¹ 0

percutaneous coronary intervention (PCI) in acute setting ¹ 0 Clopidogrel loading dose <150mg 150-300mg 300-600mg >600mg

Clinical parameter value 0 1 2 3

For instance, the logistic function may be the following

[(al * proton pump inhibitor) + (a2 * CYP2C19 metabolic status) - (a3 * ITGB3 rs5918 genotype) + (a4 * diabetes) + (a5 * left ventricular dysfunction) + (a6 * ABCB1 rsl045642 genotype) + (a7 * type C lesion) + (a8 * acute setting) - (a9 * clopidogrel loading dose) - al0] * al l .

The value of al, a2, a3, a4, a5, a6, a7, a8, a9, alO and al l, in particular of al, a2, a3, a4, a5, a6, a7, a8 and a9, may be defined based on the OR of Figure 4. Indeed, these OR of Figure 4 are indicative of the statistical weights of each parameter.

Accordingly, in the logistic function, the value of al, a2, a3, a4, a5, a6, a7, a8, a9, alO and al l are the folio wings:

al, a2, a3, a4, a5, a6, a7, a8, a9 are positive numbers, different to zero, and selected such as the ratios of the values are the folio wings;

al/a2 is 1.1 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

a3/a2 is 1 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

a4/a2 is 0.9 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

a5/a2 is 1.1 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

a6/a2 is 1.1 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

a7/a2 is 1.2 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

a8/a2 is 1.5 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

a9/a2 is 0.5 +/- 0.1, preferably 1 +/- 0.05, 0.04, 0.03 or 0.02;

and

alO is a number including zero;

al 1 is a number different to zero.

In a very particular embodiment, the logistic function is the following

(0.7825 * proton pump inhibitor) + (0.6991 * CYP2C19 metabolic status) - (0.6656 * ITGB3 rs5918 genotype) + (0.6133 * diabetes) + (0.8348 * left ventricular dysfunction) + (0.7767 * ABCB1 rsl045642 genotype) + (0.8459 * type C lesion) + (1.0896 * acute setting) - (0.3086 * clopidogrel loading dose) - 2.9262

Therefore, with this logistic function, a score less than -1.4126 is indicative of a low risk of early stent thrombosis for the subject, a score between -1.4126 to -0.4349 is indicative of an intermediate risk of early stent thrombosis for the subject, and a score of more than - 0.4349is indicative of a high risk of early stent thrombosis for the subject.

If the logistic function is amended, it is obvious for the one skilled in the art how to modify the score.

Based on the information and other clinical data of the patient, the physician will decide the most appropriate treatment with an antiplatelet therapy including a thienopyridine available on the market, in particular clopidogrel (Plavix®), prasugrel (Efient®) or ticagrelor (Brilique®). For instance, the clopidogrel treatment may be avoided in the high risk population and the treatment by prasugrel or ticagrelor might be preferred. At the opposite, the clopidogrel treatment may be preferred over the treatment by prasugrel or ticagrelor in the low risk population.

Kits and uses thereof.

The present invention also concerns a kit for determining the genotype of the SNPs rs4244285, rsl2248560, rsl045642 and rs5918. More particularly, the present invention concerns a kit for determining if a subject with a clopidogrel treatment is at risk of developing early stent thrombosis comprising means, preferably primers and/or probes, for genotyping the single nucleotide polymorphisms (SNP) rs4244285 and rsl2248560 of CYP2C19 gene, rs 1045642 of ABCB1 and rs5918 of ITGB3. The kit may comprise means, in particular products and reagents, necessary for genotyping the SNPs rs4244285, rsl2248560, rsl045642 and rs5918. Means for genotyping are well-known by the one skilled in the art. They may comprise a primer, a pair of primers, and/or a probe for each SNPs. The pair of primers is suitable for amplifying the region including the SNP. A primer may be specific of an allele of the SNP by specifically hybridizing with the sequence comprising the SNP and being capable of initiating amplification only if the allele is present, thereby the amplification being indicative of the SNP's allele. The probe hybridizes with the sequence including the SNP and is specific of one allele, thereby allowing the detection of the allele. More particularly, the kit may comprise one probe for each allele of the SNP. The probes for the two distinct alleles are preferably labeled with the different dyes, in particular fluorescent dyes. Preferably, it comprises, for the SNPs rs4244285, rsl2248560, rsl045642 and rs5918, a pair of primers and one probe for each allele of the SNP.

Methods suitable for detection of the polymorphism are well known in the art. Suitable assays include allele-specific real time PCR, 5'-nuclease assays, oligonucleotide ligase assays, allele specific oligonucleotide ligation, template-directed dye-terminator incorporation, molecular beacon allele-specific oligonucleotide assays, assays employing invasive cleavage with Flap nucleases, allele-specific hybridization (ASH), dynamic allele-specific hybridization, microarray based hybridization, allele- specific ligation, primer extension, single-base extension (SBE) assays, sequencing, pyrophosphate sequencing, real-time pyrophosphate sequencing, sequence length polymorphism analysis, restriction length fragment polymorphisms (RFLP), RFLP-PCR, single-stranded conformational polymorphism (SSCP), PCR-SSCP, ARMS-PCR, fragment sizing capillary electrophoresis, temperature gradient gel electrophoresis, denaturing high performance liquid chromatography, high resolution melting of the amplicon, heteroduplex analysis, and mass array systems. Analysis of amplified sequences may be performed using various technologies such as microchips, fluorescence polarization assays, and matrix-assisted laser desorption ionization (MALDI) mass spectrometry.

Specific commercial methodologies available for SNP genotyping include, but are not limited to, TaqMan genotyping assays and SNPIex platforms (Applied Biosystems), gel electrophoresis (Applied Biosystems), mass spectrometry (e.g., MassARRAY system from Sequenom), minisequencing methods, real-time PCR, Bio-Plex system (BioRad), CEQ and SNPstream systems (Beckman), array hybridization technology (e.g., Affymetrix GeneChip; Perlegen), BeadArray Technologies (e.g., Illumina GoldenGate and Infinium assays), array tag technology {e.g., Parallele), and endonuclease-based fluorescence hybridization technology (Invader; Third Wave).

The kit may further comprise instructions allowing the determination of the risk of early stent thrombosis for the subject.

The present invention also concerns the use of a kit for genotyping the single nucleotide polymorphisms (SNP) rs4244285 and rsl2248560 of CYP2C19 gene, rsl045642 of ABCB1 and rs5918 of ITGB3 for determining the risk of developing early stent thrombosis in a subject with a clopidogrel treatment.

Other characteristics and advantages of the invention are given in the following experimental section, which should be regarded as illustrative and not limiting the scope of the present application. EXAMPLES

Using a web-based case collection and reporting system, 369 stented patients were studied including 123 patients with definite early ST on dual antiplatelet therapy age and gender matched in a 2: 1 ratio with 246 controls. All patients were genotyped for 23 genetic variants involved in clopidogrel metabolism (CYP2C19, CYP2C9, CYP2B6, CYP3A5, POR, ABCBl, PONI), platelet receptor function (P2Y12, ITGB3), and the coagulation and fibrinolytic system (MTHFR, Factor V, Fibrinogen, Prothrombin, PAI1 and VKORC1).

CYP2C19*2 (OR _und =2.53, 95% CI [1.61-3.97], pO.0001) and ABCBl TT3435 (ORu _n d=2.01, 95% CI [1.22-3.30], p=0.006) carriers were found more frequently, and CYP2C19*17 (OR _und =0.53, 95% CI [0.31-0.88], p=0.01) and ITGB3 PIA2 (OR _und =0.50, 95% CI [0.29-0.87], p=0.01) less frequently in patients with early ST than in control patients. Acute coronary syndrom, complex lesion, left ventricular ejection fraction < 35%, high clopidogrel loading doses and use of proton pump inhibitor were the non-genetic independent correlates of early ST. The accuracy of the clinical model to discriminate between early ST and controls (AUC 0.72, 95% CI [0.66-0.77] did not differ significantly from the genetic model (AUC 0.68, 95% CI [0.62-0.73] (p=0.34), but combining both led to a significant improvement in the discriminatory power of the model (AUC 0.78, 95% CI [0.73-0.83], p=0.004). By comparison, a model based on the CYP2C19*2 allele alone present a lower accuracy (AUC 0.6234 (+/-0.03), 95% CI [0.56-0.69]). Therefore, the genetic and clinico- genetic models provided by the present invention clearly improve the available tests. Among all independent predictors of early ST, the use of high clopidogrel loading doses (OR=0.73, 95% CI[0.57-0.94], p=0.01) and proton pump inhibitors (OR=2.19, 95% CI [1.28-3.72], p=0.004) were the only modifiable factors.

Results

Characteristics of the study population

Among the 233 CAD patients with definite ST who were referred to the present study, 64 occurred after complete or partial interruption of DAPT (Figure 1). Among the remaining 169 cases, we observed 46 late or very late ST. Finally, 123 patients were identified as early ST (9 acute and 114 subacute) and matched according to age and gender with 246 controls.

Cases were more frequently found to be diabetics and have a prior history of hypertension or MI (Table 1). They also presented more frequently with impaired left ventricular function, complex lesions (type C) and ACS in comparison with controls. In addition, cases received significantly lower clopidogrel loading doses than controls and were more frequently on proton pump inhibitors. Table 1. Baseline characteristics

Antithrombotic medication during PCI

Clopidogrel loading dose <0.0001

<600mg 90(73.2%) 123 (50%)

>600mg 33 (26.8%) 123 (50%)

Proton pump inhibitor 85 (69.1%) 131 (53.2%) 0.004

GPIIb/IIIa inhibitor 49 (39.8%) 96 (39.0%) 0.88

VKA 6 (2.5%) 6 (2.5%) 1.00

* refer to the thrombotic stent

Genetic determinants of early stent thrombosis

Carriage of the CYP2C19*2 loss-of- function allele was more frequent in cases than controls (48.8% vs. 27.4% respectively, unadjusted OR=2.53 95%CI [1.61-3.97], p<0.0001) (Table 2), including a seven-fold higher prevalence of *2/*2 homozygotes (16.3% vs. 2.5%, unadjusted OR=7.73 95%CI [3.02-19.82], pO.0001). Of note, carriage of the CYP2C19*17 gain-of- function allele was significantly less frequent in cases than controls (20.4% vs. 32.7%, unadjusted OR=0.5395%CI [0.31-0.88], p=0.01). Additionally, the CYP2C19 metabolic status distribution was significantly different between cases and controls (Figure 2). No significant differences were observed when considering CYP2C9 or CYP2B6 metabolic status.

Table 2 Genetic analyses in cases and controls

7481G>A (rs2359612) 0.47

No. of patients with data 123 245

GG - n (%) 47 (38.2%) 79 (32.2%)

GA - n(%) 47 (38.2%) 108 (44.1%)

AA - n(%) 29 (23.6%) 58 (23.7%)

Of interest, cases were more frequently ABCBI 3435 TT homozygotes vs CT/CC (31.7% vs 18.8%, unadjusted OR=2.01, 95%CI [1.22-3.30], p=0.006) and less frequently carriers of the ITGB3 PIA2 polymorphism (16.4% vs. 27.8%, unadjusted OR=0.50, 95%CI [0.29-0.87], p=0.01) than controls.

None of the other genetic variants which have been shown to interfere with clopidogrel pharmacology (CYP2B6, CYP2C9, CYP3A5, POR, PON1 and P2Y12) or with thrombotic disorders (MTHFR, PAI1, Factor V, Prothrombin, Fibrinogen beta chain, VKORCl) were associated with ST.

To further evaluate the genetic determinants of early ST, a risk allele score was developed to determine whether there was a gene-dose effect in CYP2C19, ABCBI, and/or ITGB3 variant allele carriers. As shown in Figure 3, a stepwise increase in the risk of early ST according to the number of risk alleles carried by the patients has been identified.

Independent correlates of early stent thrombosis Multivariate stepwise logistic regression analyses were then performed to identify which clinical, angiographic and genetic variables best predicted the occurrence of early ST (Figure 4). Non-genetic independent correlates of early ST were ACS presentation, complex lesions (type C), left ventricular function <35%, diabetes mellitus, use of proton pump inhibitors and clopidogrel loading doses. Among the genetic factors, predicted CYP2C19 metabolic status (rapid, extensive, intermediate, poor) was a major determinant of early ST along with ITGB3 PLA2 and ABCBI 3435 TT carriage. There was a stepwise risk increase according to CYP2C19 metabolic status, the lowest being associated with RM and the highest with PM status.

Similar results were found when restricting the analyses to Caucasians, the overwhelming ethnicity in our study population. No significant interactions were identified between the genetic variants and each of the independent clinical and angiographic predictors. There were also no significant interactions between the CYP2C19, ABCBI and ITGB3 genetic variants. Performance of different models to predict early stent thrombosis

Successive multiple logistic regression models were built for the occurrence of early ST. The first model included only clinical and angiographic characteristics (clinical model), the second model included CYP2C19 metabolic status, ABCB1 (TT vs CC or TC) and ITGB3 genotypes (genetic model), and the third model included all of the above-mentioned correlates (comprehensive model). The respective equations are defined in table 3. Then the predictive performance of these models were compared by receiver operating characteristic curve analyses.

Table 3. Clinical, genetic and comprehensive algorithms

Clinical algorithm _;

Score = (0.6874 * proton pump inhibitor) + (0.6361 * diabetes) + (0.9312 * left ventricular dysfunction) + (0.8990 * type C lesion) + (1.0849 * acute setting) - (0.2978 * clopidogrel loading dose) - 2.0451

Genetic algorithm

Score = (0.7196 * CYP2C19 metabolic status) + (0.7191 * ABCB1 TT genotype) - (0.7277 * ITGB3 P1A2 genotype) - 1.5838

Clinical and genetic algorithm

Score = (0.7825 * proton pump inhibitor) + (0.6991 * CYP2C19 metabolic status) - (0.6656 * ITGB3 P1A2 genotype) + (0.6133 * diabetes) + (0.8348 * left ventricular dysfunction) + (0.7767 * ABCB1 TT genotype) + (0.8459 * type C lesion) + (1.0896 * acute setting) - (0.3086 * clopidogrel loading dose) - 2.9262

For all algorithms, CYP2C19 metabolic status and clopidogrel loading doses are coded as categorical variables as follows:

CYP2C19 rapid=0, normal = 1, intermediate = 2, slow = 3

Clopidogrel loading dose: <150mg =0; 150 to 300 mg = 1; 300 to 600mg =2; >600mg = 3. The inventors found that both the clinical and genetic models were able to significantly discriminate between early ST and controls; however, the accuracy was not statistically different between these two models (p=0.34) (Figure 5). The comprehensive model, a combination of clinical and genetic variables, had significantly greater power to discriminate early ST from matched controls (p=0.004). Moreover, it had greater sensitivity and specificity (67% and 79%, respectively) than the clinical (60%> and 70%>) and genetic (48%o and 78%) models. Subsequently, the positive likelihood ratio was the highest with the comprehensive model (3.4) than with the clinical (2.1) or genetic (2.0) models. Patients in the highest tertile of risk using the comprehensive model display a 6 fold increased risk of early ST than patients in the lowest tertile (OR 5.99; 95% CI 3.34-10.77) (Figure 6). Discussion

Stent thrombosis is a serious complication of coronary stenting but its low incidence makes it difficult to study in a comprehensive manner. Registry studies have identified a certain number of risk factors that may be or not connected to each other, but to the inventors' knowledge, no study has evaluated simultaneously all clinical, drug, angiographic, procedural and genetic factors of ST. The present multicenter study suggests that combining the genotypes of three genes related to clopidogrel metabolism and platelet function (CYP2C19, ABCBl, ITGB3) with clinical risk factors, significantly improves the ability to identify patients at-risk of early ST. The second major finding is that the only two modifiable factors of early ST are related to clopidogrel metabolism, namely clopidogrel loading dose and clopidogrel interaction with proton pump inhibitors. Taken together, these findings suggest that genetic testing is a useful addition to the clinical risk assessment of ST.

The present study adds to the understanding of the genetic profile of patients at risk of early ST while treated with clopidogrel. Of the 23 pre-selected genetic variants, four were found to be independently associated with the occurrence of early ST. These non-modifiable risk factors are highly prevalent. All variants except one have been shown to directly interfere with clopidogrel metabolism and are significantly associated with on-clopidogrel platelet reactivity. Notably and contrasting with a recent report (Bouman et al., 2010), the inventors found no significant correlation between carriage of the PON1 Q192R allele and ST. Paroxonase is a hepatic esterase recently shown to be an important factor in the transformation of the intermediate metabolite 2-oxo-clopidogrel into the active thiol metabolite. In this recent study, the PON1 Q192R variant allele was associated with both the pharmacodynamic response to clopidogrel and the occurrence of ST. However the limited number of ST cases (n=41) and the lack of any observed effect of CYP2C19*2, a contrasting finding with previous candidate gene and genome -wide association studies (Shuldiner et al, 2009), may have led to an overestimation of the effect of PON1 Q192R. The lack of association between PON1 Q192R and ST has also been recently reported elsewhere. In contrast, the CYP2C19*2 allele was independently associated with a high risk of ST and the prevalence of homozygotes was impressively high in patients with ST and seven- fold higher than in controls. The inventors also report that CYP2C19*17 gain-of- function allele carriers were at lower risk of early ST, a variant associated with higher on-treatment platelet inhibition and more bleeding complications (Sibbing et al., 2010) and less ischemic events in the CHARISMA study (Bhatt, 2009). This further strengthens the current evidence on the predominant role of CYP2C19 in clopidogrel metabolism. The inventors demonstrated in the present study a significant association between carriage of the ABCB1 3435 TT (vs. CT or CC) genotype and early ST. This gene encodes a drug efflux transporter, P-glycoprotein, which modulates clopidogrel absorption. It has been previously associated with reduced clopidogrel response (Mega et al, 2010b; Taubert et al, 2006) but with variable clinical consequences (Simon et al, 2009; Mega et al, 2010b; Wallentin et al, 2010)). One explanation for the discrepancy may be that ABCB1 C3435T is a synonymous variant (p.llel 145Ile) that affects protein conformation and substrate specificity (Kimchi-Sarfaty et al, 2007) which may vary according to clopidogrel dosing strategies as recently shown by a greater impact of this variant at the time of clopidogrel loading versus during the maintenance phase (Campo et al, 2011). The present finding that ABCB1 C3435T carriage and low clopidogrel loading dose were independent correlates of early ST and the lack of saturable clopidogrel intestinal absorption as demonstrated by previous investigation looking at the effect of different loading dose strategies on clopidogrel response (Collet et al, 2008; L'Allier et al, 2008; Bonello et al, 2010b) are not supportive of such hypothesis.

Of note, assessment of both ABCB1 C3435T and CYP2C19 shows that the two genes offer complementary information and nearly ¾ of patients with ST had an increased risk genetic profile define as the number of alleles at risk. The inventors' regression models indicate also that in patients with a low clinical likelihood of ST, the weight of genetic factors was major to predict ST.

The ITGB3 gene encodes for the integrin β3, a component of the GPIIb/IIIa platelet receptor which mediates the final pathway of platelet aggregation. It is highly polymorphic with two common allelic isoforms, namely PLA1 and PLA2. The relations of PLA2 (p.Leu59Pro) with the occurrence of acute coronary syndrome but also with the biological response to aspirin or GPIIbllla receptor antagonists have been controversial (Pellitero et al, 2010; Weiss et al, 1996; Zhu et al, 2000). The present investigation shows that the PLA2 polymorphism was less frequent in ST patients than in control patients without ST. This information appears to be new and may add to the accuracy of our predictive model.

The overlap between the clinical and genetic regression models appears limited and no significant interaction was found. Obviously, the genetic information cannot be captured by the clinical or procedural characteristics. The lack of variation in the OR values of early ST clinical predictors before and after consideration of the genetic correlates and their close absolute values with that of clinical and angiographic correlates further suggest a similar extent of the predicted risk and a lack of correlation. Materials and Methods

Selection of Patients

Patients >18 years of age with a previous angio graphically documented early (<30 days) ST according to the Academic Research Consortium (ARC) definitions (Cutlip et al, 2007) were enrolled. Angiographic definite ST consisted of partial or complete occlusion within a previously implanted stent and evidence of fresh thrombus (in the stent or in the 5 mm proximal or distal to the stent) associated with at least one of the following criteria: ischemic symptoms, ischemic ECG changes or elevation of biomarkers. Among early ST, acute ST occurred within 24 hours after stent implantation and subacute occurred between >24 H and < 30 days after stent implantation. Control subjects were recruited in the same participating centers and had to be on DAPT (aspirin and clopidogrel), followed-up for at least one year without history of ST. For the case-control comparison, a ratio of one ST case for two age and gender matched controls was used.

Baseline clinical, angiographic and procedural characteristics of case and control subjects were carefully collected. Coronary angiograms of both case and control subjects were reviewed by two independent interventional cardiologists. The status of antiplatelet therapy at the time of ST was also carefully evaluated. Definite ST cases that occurred after partial or complete interruption of DAPT were excluded from the present analysis.

Study objectives

The primary objective was to identify independent clinical, angiographic and genetic determinants of early ST. The secondary objective was to develop a predictive model of early ST.

Genotyping

Genomic DNA of case and control subjects was extracted from peripheral-blood leukocytes by standard procedures (Puregene DNA isolation kit, Merck Eurolab, Lyon, France). Genotyping was performed using TaqMan Validated SNP assays with the 7900HT sequence Detection System (Applied Biosystems, Courtaboeuf, France). All participants were genotyped for the common CYP2C19 *2 (rs4244285) loss of-function allele and CYP2C19*3 (rs4986893), CYP2C19H (rs28399504), CYP2C19*5 (rs56337013), CYP2C19*6 (rs72552267), CYP2C19*17 (rsl2248560), CYP2C9*2 (rsl799853), CYP2C9*3 (rsl057910), CYP2B6*5 (rs3211371), CYP2B6*9 (rs3745274), CYP3A5*3 (rs776746), POR*28 (rsl057868), PON1 Q192R (rs662), PON1 L55M (rs854560), ABCB1 C3435T (rs 1045642), P2Y12 (rs2046934), ITGB3 (rs5918), MTHFR (rsl801133), PAIl (rsl799889), Factor V (rs6025), Prothrombine G20210A (rs 1799963), Fibrinogen beta chain (rs 1800790) and VK0RC1 (rs2359612), as previously described in Collet et al, 2009.

CYP2C19, CYP2C9, and CYP2B6 metabolism status was determined for cases and controls based on identified polymorphisms and their effect on enzymatic function (Ingelman- Sundberg et al, 2007; Mega et al, 2009)

Table 4: predicted Metabolic phenotype according to observed genotype

PM poor metabolizer, IM intermediate metabolizer, EM extensive metabolizer, RM Rapid metabolizer.

Statistical analysis

All analyses were performed with the use of SAS software, version 9.1 (SAS Institute, USA). All single-nucleotide polymorphisms (SNPs) were tested for deviation from Hardy- Weinberg equilibrium with the chi-square test. Continuous variables were expressed as meantSD unless otherwise stated and categorical variables as frequencies and percentages. Baseline characteristics of patients by genotypes were compared using the χ ² test for categorical variables and Student's t-test or one-way analysis-of- variance for continuous variables as appropriate. Comparisons are expressed as univariate hazard ratios and 95% CIs including the entire duration of follow-up.

A univariate logistic model was used to compare all the characteristics (clinical, angiographic and genetic). Stepwise multivariable logistic regression analysis was used to identify independent variables associated with the occurrence of early ST. The models included baseline demographic (age, sex, body-mass index, smoking, ancestry), clinical (hypertension, diabetes mellitus, previous MI, creatinine clearance below 60ml/min, left ventricular ejection fraction<45%>, clopidogrel loading dose, use of proton pump inhibitor), and angiographic (ACC/AHA type C lesion, number of vessel with disease, indication of PCI, dissection post-PCI) characteristics, predicted metabolic status (CYP2C19, CYP2C9 and CYP2B6) and genetic variants (CYP3A5*3, POR*28, PON1 Q192R and L55M, ABCB1 C3435T, P2Y12 H1/H2, ITGB3 T1565C, MTHFR C677T, ΡΑΠ 5G/4G, Factor V G1691A, Prothrombin G20210A, Fibrinogen G-455A and VKORCl). Loading doses were classified as : no loading dose, 150-300mg, 300-600mg, more than 600mg. CYP2C19 metabolic status was defined as: rapid (RM), extensive (EM), intermediate (IM) and poor metabolizer (PM) (table 4). The regression parameters of the independent variables were used to derive the weighted equations.

The main analyses were age and sex adjusted. Additional analyses were also restricted to the predominant ethnicity.

Nonparametric receiver operating characteristic (ROC) curves were used to assess the discriminatory power of three prediction algorithms to distinguish cases and controls, a clinical/angiographic model, a genetic model and a comprehensive model which included all the correlates of early ST. Pairwise comparisons of AUC were performed according to Delong et al. (1988). The optimal cut-off point was calculated to provide the greatest sum of sensitivity and specificity. Odds ratios were calculated to estimate the risk of early ST according to each model.

REFERENCES

Bhatt DL. Transcatheter Cardiovascular Therapeutics meeting (TCT 2009), San Francisco, CA, September 24, 2009.

Bonello et al. J Am Coll Cardiol. 2010a;56(2): l 12-116.

Bonello et al. J Am Coll Cardiol. 2010b;56:1630-1636.

Bouman et al. Nat Med. 2010;(1): 110-116.

Campo et al. J Am Coll Cardiol. 2011;57:2474-2483.

Collet et al. Circulation. 2008;118: 1225-1233.

Collet et al. Lancet. 2009;373: 1172-1173.

Cutlip et al. Circulation. 2007;115:2344-2351.

DeLong et al. Biometrics. Sep 1988;44(3):837-845.

Hulot et al. J Am Coll Cardiol 2010;56(2): 134-143.

Ingelman-Sundberg et al. Pharmacol Ther. Dec 2007;116(3):496-526.

Kazui et al. Drug Metab Dispos. 2009;38(l):92-99.

Kimchi-Sarfaty et al. Science. 2007;315:525-528.

Klerk et al. Jama. Oct 23-30 2002;288(16):2023-2031.

LAllier et al. J Am Coll Cardiol. . 2008;51(11): 1066-1072. Mega et al. N Engl J Med. Jan 22 2009;360(4):354-362.

Mega et al. JAMA. 2010a;304(16): 1821-1830.

Mega et al. Lancet. 2010b;376(9749): 1312-1319.

Pellitero et al. Thromb Haemost. Mar 2010;103(3):630-637. Satra et al. Pharmacogenomics. Feb 201 1 ;12(2): 195-203.

Shuldiner et al. JAMA. Aug 26 2009;302(8):849-857.

Sibbing et al. Circulation. 2010;121(4):512-518.

Simon et al. N Engl J Med. 2009;360:363-375.

Taubert et al. Clin Pharmacol Ther. 2006;80:486-501.

Wallentin et al. Lancet. 2010 376: 1320-1328.

Wang et al. Circulation. Mar 28 2006; 1 13(12): 1615-1621. Weiss et al. N Engl J Med. Apr 25 1996;334(17):1090-1094. Zhu et al. Am J Cardiol. Nov 1 2000;86(9): 1000-1005, A1008.