Increased risk of atherothrombotic events associated with cytochrome P450 3A5 polymorphism in patients taking clopidogrel

Phase 1

The study protocol was approved by the Institutional Review Board of Seoul National University Hospital (SNUH), and the study was conducted at the SNUH Clinical Trial Center. All subjects provided written informed consent before enrolment.

Among 486 unrelated people who had undergone genotyping for the CYP3A5 *3 allele to establish baseline data for a DNA bank for the study of pharmacogenetics in Koreans (data not published), 16 people with expressor alleles (*1*1 or *1*3 alleles) and 16 with non-expressor alleles (*3*3 alleles) volunteered to participate in our study. The 2 groups did not differ significantly in terms of age, height or weight. The participants were ascertained to be healthy by medical history, physical examination, vital signs and routine clinical laboratory tests performed within 3 weeks before the start of the study. None was a regular heavy drinker or a smoker, or had a body weight more than 120% of his or her ideal weight.

Patients were given a loading oral dose of 300 mg of clopidogrel (Plavix; Sanofi-Synthelabo Korea) followed by daily oral doses of 75 mg for 6 days. A platelet aggregation test was performed at baseline and at 4 hours, 24 hours and 6 days after the loading dose. After a washout period of 28 days, we performed the same experiment but with a pretreatment of 200 mg of itraconazole (Sporanox cap; Janssen Korea) for 4 days to assess the effect of CYP3A5 genotype on the antiplatelet efficacy of clopidogrel. Itraconazole is a well-known CYP3A inhibitor and has been frequently used in pharmacokinetic and pharmacodynamic experiments of CYP3A4 substrates.20^–22 Therefore, we chose itraconazole for CYP3A4 inhibition. (The 2 dosing regimens for phase 1 of the study are appear in Appendix 1, available online at www.cmaj.ca/cgi/content/full/174/12/1715/DC1.) Participants were instructed to refrain from taking other medications, herbal drugs, alcohol, caffeinated beverages and grapefruit products for the 7 days before and the 7 days during the clopidogrel-only regimen and the 10 days before and the 7 days during the clopidogrel plus itraconazole regimen.

Platelet aggregation was measured using platelet-rich plasma stimulated with 5 μmol/L ADP and was assessed with a Chronolog Lumi-Aggregometer (model 560-Ca; Chronolog Corporation) equipped with an AggroLink software package.23 After inverting the vacutainer tube 3–5 times, the blood-citrate mixture was centrifuged at 1200 g for 2.5 minutes. The platelet count was determined in the platelet-rich plasma and adjusted to 3.0 × 10⁵/mL with homologous platelet-poor plasma. Aggregation was expressed as the maximal percentage change in light transmittance from the baseline value, using platelet-poor plasma as a reference. The change in platelet aggregation after clopidogrel treatment was expressed as the absolute reduction in aggregation achieved by 5 μmol/L of ADP after clopidogrel administration compared with the baseline aggregation values before clopidogrel administration.6

Phase II

The study protocol was approved by the SNUH Institutional Review Board, and the study was conducted at the SNUH cardiovascular centre. All subjects provided written informed consent before enrolment.

We enrolled consecutive patients at the Seoul National University Cardiovascular Center who had undergone successful coronary angioplasty with implantation of bare-metal stents after pretreatment with clopidogrel for at least 24 hours from November 2001 to August 2003. We excluded patients who underwent primary percutaneous coronary intervention, received drug-eluting stents, were treated with ticlopidine or were in cardiogenic shock. DNA analysis was performed in all cases to determine the patients' CYP3A5 genotypes. All patients gave informed consent to participate in the study.

Coronary angioplasty and stent implantation were performed using conventional techniques.24 Antiplatelet therapy consisted of ASA (100–300 mg/d, prescribed indefinitely) and clopidogrel (75 mg/d after a 300-mg loading dose), administered for at least 4 weeks after the procedure. Other anti-anginal and antihypertensive medications were given at the treating physician's discretion. All treating physicians were blind to the results of the DNA analysis.

Genomic DNA was prepared from the whole blood of each subject by means of standard methods.25 The presence of the CYP3A5*3 allele was determined by mismatch polymerase chain reaction (PCR)-restriction fragment length polymorphism analysis, as previously reported.20 In brief, genomic DNA was amplified using CYP3A5 6956Fm (5′-CTT TAA AGA GCT CTT TTG TCT CTC A-3′) as the forward primer and CYP3A5 7155R (5′-CCA GGA AGC CAG ACT TTG AT-3′) as the reverse primer. After amplification, the 200-base pair (bp) PCR product was digested using the enzyme DdeI. Digestion of the CYP3A5*3 allele yielded fragments of 22, 71 and 107 bp, and digestion of the CYP3A5*1 allele yielded fragments of 71 and 129 bp.

The end point of phase 2 was a composite of atherothrombotic events (cardiac death, myocardial infarction and nonhemorrhagic cerebral infarction) within 1 and 6 months after stent implantation. The cause of all deaths was regarded as cardiovascular unless there was documented evidence of a clear noncardiovascular cause. Myocardial infarction was defined as the presence of at least 2 of the following findings: ischemic symptoms; cardiac enzyme (CK MB) concentration at least twice the upper limit of normal; or new electrocardiographic changes compatible with myocardial infarction. Nonhemorrhagic cerebral infarction was defined as a new focal neurologic deficit of vascular origin lasting at least 24 hours that was proven to be nonhemorrhagic by either CT or MRI scanning. The follow-up protocol included a medical visit at the outpatient clinic at 1 month after stent placement and every 2 or 3 months thereafter. Clinical events were assessed on the basis of the information provided by hospital readmission records, the referring physician or a phone interview with the patient. The investigators who evaluated the clinical end points were blinded to the results of the DNA analysis.

Statistical analysis

Continuous variables are presented as means and standard deviations (SDs). In phase 1, a repeated-measures analysis of variance was used to compare platelet aggregation test results at baseline, 4 hours, 24 hours and 6 days after clopidogrel administration, and the Student's t test was used to compare the degree of platelet inhibition between the 2 genotype groups.

In phase 2, the χ² test was used to compare the incidence of clinical end points between the genotype groups. A logistic regression analysis, which included age, sex, diabetes mellitus, hypertension, hypercholesterolemia, previous myocardial infarction, clinical diagnosis, CYP3A5 genotype, left ventricular ejection fraction, stent size, multivessel disease, multivessel intervention and the number of co-administered CYP3A substrates or inhibitors (lipophilic statins, metoprolol, diltiazem, nifedipine, cimetidine and losartan) as covariates, was performed to identify determinants of atherothrombotic events (cardiovascular death, myocardial infarction and nonhemorrhagic stroke) after coronary artery stenting within 6 months. A p value below 0.05 was considered statistically significant.

Phase 1

The inhibition of platelet aggregation after clopidogrel treatment was significant at all points compared with baseline in both of the genotype groups (Fig. 3). When we compared the efficacy of clopidogrel in terms of the change in aggregation from baseline, there were no significant differences between the 2 groups, although we observed a slight tendency for those with the CYP3A5 expressor genotype to achieve greater inhibition of platelet aggregation (Fig. 4).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3: Platelet aggregation in healthy volunteers taking clopidogrel, by CYP3A5 genotype (non-expressor v. expressor).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4: Change in platelet aggregation from baseline after administration of clopidogrel, by CYP3A5 genotype.

After the washout period and treatment with the CYP3A4 inhibitor itraconazole, clopidogrel significantly inhibited platelet aggregation at all time points only among the subjects with the CYP3A5 expressor genotype (Fig. 5). In contrast, clopidogrel had a significant effect only on day 7 among those with the non-expressor genotype (Fig. 5, right panel). In terms of the change in aggregation from baseline, the change was significantly greater among those with the expressor genotype than among those with the non-expressor genotype at 4 hours, 24 hours and 6 days after baseline (Fig. 6).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 5: Platelet aggregation in healthy volunteers taking clopidogrel after pretreatment with the CYP3A inhibitor itraconazole. NS = not significant.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 6: Change in platelet aggregation from baseline after administration of clopidogrel and pretreatment with itraconazole.

Phase 2

A total of 348 patients were enrolled in phase 2. In these patients, 453 lesions were treated with angioplasty and implantation of bare-metal stents. Genotype analysis revealed the following frequency among the patients: 33 (9%) had the CYP3A5 *1*1 allele, 122 (35%) had the *1*3 allele, and 193 (55%) had the *3*3 allele; which gave a frequency of non-expressor (*3) allele of 73%. These findings were consistent with the Hardy–Weinberg expectations. The patients were divided into 2 groups: those with the CYP3A5 expressor alleles (n = 155) and those with the non-expressor allele (n = 193). The clinical and angiographic characteristics of the patients did not differ significantly between the 2 groups (Table 1), nor was there a significant difference in the duration of clopidogrel treatment (4.1 [SD 2.3] and 3.9 [SD 2.3] months in the non-expressor and expressor groups respectively) or co-administered drugs (Table 2).

Six-month clinical follow-up data were available for all patients (Table 3). During this period, atherothrombotic events occurred in 17 patients (14 in the non-expressor group and 3 in the expressor group). One myocardial infarction in the non-expressor group was due to subacute stent thrombosis; all cases of stroke were confirmed as nonhemorrhagic by means of imaging studies. All of the patients in the non-expressor group who experienced an atherothombotic event were taking both clopidogrel and ASA at the time.

Multivariable analysis revealed the following as independent predictors of atherothrombotic events within 6 months after coronary angioplasty and bare-metal stent implantation: CYP3A5 non-expressor genotype, every increase in the number of co-administered CYP3A substrates or inhibitors, and a cholesterol level of 5.18 mmol/L (200 mg/dL) or higher (Table 4).

In the subgroup analysis according to genotype, the number of CYP3A metabolizers was an independent predictor of atherothrombotic events only in patients with the CYP3A5 non-expressor genotype but not in those with the expressor genotype (Table 5).