European Journal of Clinical Pharmacology

, Volume 64, Issue 9, pp 877–882

CYP3A4*1G polymorphism is associated with lipid-lowering efficacy of atorvastatin but not of simvastatin

Authors

  • Yuan Gao
    • Department of PharmacologySchool of Medicine, Zhengzhou University
    • Department of PharmacologySchool of Medicine, Zhengzhou University
  • Qiang Fu
    • Department of PharmacologySchool of Medicine, Zhengzhou University
Pharmacogenetics

DOI: 10.1007/s00228-008-0502-x

Cite this article as:
Gao, Y., Zhang, L. & Fu, Q. Eur J Clin Pharmacol (2008) 64: 877. doi:10.1007/s00228-008-0502-x

Abstract

Purpose

Our aim was to observe the impact of CYP3A4*1G genetic polymorphism on lipid-lowering efficacy of statins.

Methods

We studied 217 unrelated hyperlipidemic patients who prospectively received atorvastatin and 199 patients who received simvastatin as a single-agent therapy (20 mg day-1 p.o.) for 4 weeks. Genotyping of CYP3A4*1G was conducted by a polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP) analysis. Serum triglyceride (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) levels were determined before and after treatment by enzymatic assays.

Results

The frequency of CYP3A4*1G in Chinese hyperlipidemic patients was 0.276. After atorvastatin treatment, the mean percentage reduction in serum TC was 16.8 ± 3.3% (*1/*1), 17.8 ± 3.8% (*1/*1G), and 20.9 ± 5.0% (*1G/*1G), respectively. The CYP3A4*1G polymorphism had a gene-dose-dependent effect on percentage reduction in serum TC (P < 0.01). Conversely, there was no significant association between lipid-lowering efficacy of simvastatin and CYP3A4*1G polymorphism.

Conclusions

Carrying CYP3A4*1G increase the lipid-lowering efficacy of atorvastatin and may have no significant effect on simvastatin treatment.

Keywords

CYP3A4PolymorphismsSimvastatinAtorvastatinLipid-lowering efficacy

Introduction

Simvastatin and atorvastatin are both 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors widely used for treating hypercholesterolemia and hypertriglyceridemia [1]. However, considerable variations in interindividual lipid-lowering responses to these drugs in hypercholesterolemic subjects have been reported in clinical trials. A deeper understanding of the mechanisms mediating attenuated response or treatment failure [2] associated with statin therapy in different human populations is essential to individualize preventive antilipidemic interventions and reduce the global public health burden of coronary heart disease.

As the most abundant group of CYPs in the liver and small intestine, CYP3A4 strongly affect the oral bioavailability and clearance of many drugs, and it is estimated that CYP3A4 is involved in the metabolism of more than half of all commonly used drugs [36]. Genetic variation within the CYP3A4 gene may contribute to interindividual variability in drug metabolism. It has been suggested that approximately 90% of interindividual differences in hepatic CYP3A4 activity is due to genetic variation [7]. Single nucleotide polymorphisms (SNPs) are the most common form of genetic variation in the CYP3A4. To date, more than 39 SNPs of the CYP3A4 gene have been published on the Human Cytochrome P450 Allele Nomenclature Committee homepage (http://www.imm.ki.se/CYPalleles). Among the CYP3A4 variants, alleles with nonsynonymous SNPs, i.e., CYP3A4*2, *4, *5, *6, *17, and *18, have been shown to alter enzyme activity compared with the wild type [8]. Several studies have reported the association between the lipid-lowering response to simvastatin or atorvastatin and genetic polymorphisms of CYP3A4. Wang et al. [9] found that the mean percentage reductions in total cholesterol (TC) and triglycerides (TG) in CYP3A4*1/*4 subjects were higher than those in CYP3A4*1/*1 subjects after simvastatin treatment. Kajinami et al. [10] found the CYP3A4*1B variant was associated with higher posttreatment low-density lipoprotein cholesterol (LDL-C), whereas the CYP3A4*3 variant was associated with lower levels of LDL-C after atorvastatin treatment. However, Fiegenbaum et al. [11] did not find the association between the CYP3A4*1B variant and the efficacy of simvastatin.

More recently, CYP3A4*1G in intron 10 of the CYP3A4 gene was found by direct sequencing, initially in Japanese, which led to G to A substitution at position 82266 (GenBank No.280107.1) [12]. The frequency of the CYP3A4*1G variant allele was 0.249 in a Japanese population [12] and 0.221 in a Chinese population [13]. Although CYP3A4*1G is a high-frequency allele in Asians, the alteration of function remains unclear in vitro and in vivo. Moreover, there have been few in vivo clinical reports associated with the CYP3A4*1G allele. A recent study showed that the CYP3A4*18B variant (actually called CYP3A4*1G but not CYP3A4*18B in the study, because the last variant contains a Leu293Pro change besides the 20230 G>A substitution) affected cyclosporine pharmacokinetics, probably resulting from a higher enzymatic activity of this mutation in healthy subjects [14]. In the case of simvastatin and atorvastatin, CYP3A4 plays a central role in their metabolism [15]. We hypothesized that the CYP3A4*1G variant could contribute to the variable response of simvastatin or atorvastatin therapy. To test this hypothesis, we performed a prospective study to observe the impact of CYP3A4*1G genetic polymorphism on lipid-lowering efficacy of statins (simvastatin and atorvastatin).

Materials and methods

Subjects

A group of 423 unrelated hyperlipidemic patients (atorvastatin n = 221, male 92 and female 129; simvastatin n = 202, male 103 and female 99) (TG >1.78 mmol/l or TC >6.00 mmol/l or LDL-C >3.36 mmol/l) were recruited from hospitalized and nonhospitalized patients living in the Henan province of China. All subjects were of Chinese Han nationality. For this prospective study, all patients were screened by physical examination, medical history, and clinical laboratory evaluation and were invited to participate in this study. Exclusion criteria included the following: unstable or uncontrolled clinically significant disease, uncontrolled hypothyroidism or diabetes, and impaired hepatic or renal function. They were asked not to drink caffeine-containing beverages and to refrain from alcohol intake for at least 7 days before and after taking the study medication. None of the patients used any lipid-lowering drug or inducers or inhibitors of CYP3A for at least 2 weeks before entry into this study. Subjects entering the study were treated with 20 mg simvastatin or atorvastatin daily for 4 weeks. All patients gave written informed consent, and the study protocol was approved by the Ethics Committee of Zhengzhou University.

Genomic DNA extraction

Venous blood samples (2 ml) for CYP3A4 genotyping were obtained from all patients. DNA was extracted from leukocytes using a standard phenol/chloroform procedure [16].

Genotyping assays

Genotyping of the CYP3A4*1G allele was conducted by polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). Briefly, CYP3A4*1G (G20230A) was amplified using the forward primer 5′-CACCCTGATGTCCAGCAGAAACT-3′ and reverse primer 5′-AATAGAAAGCAGATGAACCAGAGCC-3′ [Takara Biotechnology (Dalian) Co. Ltd, China]. PCR reactions were performed in a total volume of 25 μl comprising 2.5 μl of 10 × PCR buffer, 50 pM of each primer, 0.2 mM of each deoxynucleotide triphosphate (dNTP), 2 U Taq deoxynucleic acid (DNA) polymerse (Takara Biotechnology), and 100 ng of genomic DNA as a template. Amplification conditions involved an initial denaturation for 7 min at 94℃, followed by 35 cycles of 30 s at 94℃, 1 min at 62℃, and 1 min at 72℃, with a 5-min final extension at 72℃. After PCR amplification, 8 μl of PCR product of 287 bp was digested for 4 h at 37℃ with 10 U Rsa I (Takara Biotechnology). The digested PCR products were analyzed by electrophoresis on a 2% agarose gel in the presence of ethidium bromide. The bands of DNA fragments were visualized under ultraviolet light. Rsa I digestion of the CYP3A4*1G allele gave a band of 287 bp. The wild-type DNA created one restriction site, and Rsa I digestion gave 217- and 70-bp fragments. The results for each genotype were confirmed in randomly selected individuals by direct sequence analysis.

Evaluation of lipid-lowering efficacy of simvastatin and atorvastatin

The hyperlipidemic patients were treated with 20 mg simvastatin or atorvastatin daily for 4 weeks. TG, TC, LDL-C, and high-density lipoprotein cholesterol (HDL-C) levels were determined by enzymatic assays before and after taking the study medication for 4 weeks (Roche 800, Japan). A valid lipid assessment required that measurements were made on a 12-h fasting blood draw. The primary effects measured at week 4 were the mean percentage change in TG, TC, LDL-C, and HDL-C from baseline.

Statistical analysis

Statistical analysis was conducted using the package SPSS 12.0 (SPSS Inc, Chicago, IL, USA) for Windows, and a two tailed p-value of  < 0.05 was deemed statistically significant. Data were presented as mean ± standard deviation (SD). Data were compiled according to genotype, and allele frequencies were estimated from the observed numbers of each specific allele. Chi-square test was used to verify Hardy–Weinberg equilibrium. Interethnic differences in allelic frequencies were analyzed by chi-square test or Fisher’s exact test of probabilities. To test the effects of genotype, comparisons by one-way analysis of variance (ANOVA) with post hoc Bonferroni correction for multiple comparisons was performed before and after adjusted for age, gender, and pretreatment TG, TC, LDL-C, and HDL-C levels.

Results

Genetic polymorphism in the CYP3A4 gene in Chinese

Genotyping data were taken from 416 Chinese hyperlipidemic patients. Seven subjects could not be genotyping satisfactorily. Among the 416 subjects, there were 211 wild-type homozygotes (*1/*1), 180 heterozygotes (*1/*1G), and 25 mutant homozygotes (*1G/*1G) for CYP3A4*1G (Fig. 1). The allele frequency of CYP3A4*1G was 0.276 in Chinese hyperlipidemic patients. The allele frequency was in Hardy–Weinberg equilibrium.
https://static-content.springer.com/image/art%3A10.1007%2Fs00228-008-0502-x/MediaObjects/228_2008_502_Fig1_HTML.gif
Fig. 1

Detection of polymorphism in CYP3A4*1G using polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). Following PCR and detection of G20230A by Rsa I digestion, oligonucleotides were analyzed by electrophoresis on a 2% agarose gel in the presence of ethidium bromide. Lane L is 50-bp DNA marker; lanes 1, 3 are from two samples of CYP3A4*1/*1G (w/m); lane 2 is from a sample of CYP3A4*1G/*1G (m/m); lane 4 is a sample of CYP3A4*1/*1 (w/w)

Association between CYP3A4*1G genetic polymorphism and lipid-lowering efficacy

Demographic and clinical characteristics of the patient population according to genotypes are given in Table 1 for the atorvastatin treatment panel and Table 2 for the simvastatin treatment panel.
Table 1

Association between the CYP3A4 polymorphism and serum lipid and lipoprotein levels after atorvastatin treatment

 

All subjects

*1/*1

*1/*1G

*1G/*1G

P value

ANOVA

*1/*1 vs. *1/*1G

*1/*1 vs. *1G/*1G

*1G/*1G vs. *1/*1G

No. of subjects

217

107

99

11

    

Age (years)

53.0 ± 7.9

52.5 ± 8.1

53.3 ± 7.4

53.6 ± 7.2

0.764

0.827

0.765

0.623

BMI (kg/m2)

27.0 ± 2.6

26.5 ± 3.2

27.3 ± 2.4

27.1 ± 2.6

0.688

0.713

0.648

0.368

Baseline (mmol/L)

TC

7.09 ± 0.48

7.06 ± 0.47

7.10 ± 0.48

7.14 ± 0.53

0.817

0.630

0.612

0.777

LDL-C

3.62 ± 0.49

3.61 ± 0.48

3.60 ± 0.48

3.75 ± 0.50

0.638

0.948

0.364

0.351

HDL-C

1.67 ± 0.29

1.61 ± 0.29

1.73 ± 0.27

1.52 ± 0.25

0.174

0.161

0.268

0.079

TG

1.93 ± 0.37

1.96 ± 0.44

1.91 ± 0.28

1.84 ± 0.26

0.486

0.410

0.303

0.518

Treatment (mmol/L)

TC

5.85 ± 0.45

5.88 ± 0.30

5.83 ± 0.39

5.65 ± 0.67

0.248

0.438

0.107

0.212

LDL-C

2.64 ± 0.45

2.63 ± 0.43

2.63 ± 0.42

2.72 ± 0.53

0.789

0.926

0.493

0.523

HDL-C

1.65 ± 0.28

1.62 ± 0.28

1.68 ± 0.26

1.55 ± 0.18

0.191

0.159

0.411

0.144

TG

1.48 ± 0.24

1.49 ± 0.30

1.47 ± 0.17

1.41 ± 0.19

0.552

0.633

0.290

0.404

Percent changea

TC

−17.4 ± 3.7

−16.8 ± 3.3

−17.8 ± 3.8

−20.9 ± 5.0

0.001

0.031

0.000

0.011

LDL-C

−27.2 ± 5.6

−27.3 ± 5.5

−27.0 ± 5.7

−27.9 ± 5.1

0.993

0.926

0.938

0.975

HDL-C

0.2 ± 9.3

1.2 ± 10.1

−2.4 ± 7.8

3.6 ± 10.4

0.074

0.055

0.405

0.090

TG

−22.9 ± 8.7

−23.3 ± 8.4

−22.3 ± 9.1

−22.1 ± 7.6

0.918

0.725

0.888

0.765

All data are presented as mean ± standard deviation (SD)

BMI body mass index, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, TG triglyceride, ANOVA analysis of variance

aP value are adjusted by age, gender, and pretreatment TG, TC, LDL-C and HDL-C levels

Table 2

Association between the CYP3A4 polymorphism and serum lipid and lipoprotein levels after simvastatin treatment

 

All subjects

*1/*1

*1/*1G

*1G/*1G

P value

ANOVA

*1/*1 vs. *1/*1G

*1/*1 vs. *1G/*1G

*1G/*1G vs. *1/*1G

No. of subjects

199

104

81

14

    

Age (years)

53.8 ± 8.1

53.8 ± 8.0

53.9 ± 8.4

55.0 ± 7.9

0.872

0.917

0.601

0.645

BMI (kg/m2)

27.4 ± 3.4

27.3 ± 3.4

27.5 ± 3.6

27.8 ± 3.1

0.884

0.923

0.626

0.657

Baseline (mmol/L)

TC

7.12 ± 0.51

7.08 ± 0.45

7.14 ± 0.51

7.36 ± 0.84

0.138

0.407

0.057

0.134

LDL-C

3.59 ± 0.48

3.58 ± 0.47

3.58 ± 0.47

3.71 ± 0.67

0.641

0.949

0.368

0.359

HDL-C

1.64 ± 0.35

1.63 ± 0.36

1.66 ± 0.35

1.55 ± 0.24

0.564

0.616

0.420

0.294

TG

2.00 ± 0.58

2.01 ± 0.69

1.99 ± 0.41

1.99 ± 0.51

0.996

0.935

0.948

0.982

Treatment (mmol/L)

TC

5.81 ± 0.43

5.78 ± 0.40

5.82 ± 0.45

5.90 ± 0.48

0.591

0.569

0.338

0.514

LDL-C

2.60 ± 0.44

2.59 ± 0.44

2.60 ± 0.42

2.72 ± 0.55

0.574

0.878

0.294

0.341

HDL-C

1.64 ± 0.26

1.64 ± 0.27

1.63 ± 0.27

1.64 ± 0.18

0.965

0.791

0.929

0.962

TG

1.46 ± 0.40

1.47 ± 0.50

1.47 ± 0.24

1.39 ± 0.30

0.775

0.998

0.488

0.494

Percent changea

TC

−18.4 ± 3.7

−18.2 ± 3.5

−18.4 ± 3.7

−19.5 ± 5.3

0.961

0.858

0.867

0.796

LDL-C

−27.7 ± 5.7

−27.9 ± 5.7

−27.5 ± 6.0

−26.8 ± 3.7

0.908

0.734

0.722

0.857

HDL-C

1.8 ± 11.9

2.6 ± 12.2

0.1 ± 11.2

6.7 ± 12.8

0.090

0.080

0.375

0.077

TG

−25.6 ± 10.5

−25.4 ± 11.3

−25.1 ± 9.8

−29.1 ± 9.2

0.499

0.758

0.305

0.239

All data are presented as mean ± standard deviation (SD)

BMI body mass index, TC total cholesterol, LDL-C low-density lipoprotein cholesterol, HDL-C high-density lipoprotein cholesterol, TG triglyceride, ANOVA analysis of variance

aP value adjusted by age, gender, and pretreatment TG, TC, LDL-C, and HDL-C levels.

After treatment with 20 mg atorvastatin or simvastatin daily for 4 weeks, TG, TC, and LDL-C concentrations were decreased from baseline, on average, by 22.9 ± 8.7% or 25.6 ± 10.5%, 17.4 ± 3.7% or 18.4 ± 3.7%, and 27.2 ± 5.6% or 27.7 ± 5.7%, respectively (Tables 1 and  2) (P <  0.001). HDL was not significantly affected after atorvastatin or simvastatin treatment.

We found a significant influence of CYP3A4 genotype on percentage reduction in TC from baseline after atorvastatin treatment. The mean percentage reduction in serum TC was 16.8 ± 3.3% (*1/*1), 17.8 ± 3.8% (*1/*1G), and 20.9 ± 5.0% (*1G/*1G), respectively. Statistical difference was observed between the *1/*1 and *1G/*1G groups, between the *1/*1G and *1G/*1G groups, and between the *1/*1 and *1/*G groups. The CYP3A4*G polymorphism had a dose-dependent effect on percentage reduction in serum TC (P < 0.01) (Table 1). On the other hand, the percentage reduction of LDL-C and TG did not show a significant difference between the three genotypes. Multiple regression analysis using the percentage change in TC as dependent variable and age; gender; pretreatment TG, TC, LDL-C, and HDL-C; levels and genotype group as independent variables demonstrated that genotype group (P = 0.0006) and age (P = 0.008) were significant determinants of the percentage change in TC. Regression models were also applied to predict the variables involved in LDL-C response. Age (P = 0.027) and pretreatment HDL-C level (P = 0.029) entered the model. The percentage change in TG was significantly related only to pretreatment TG level (P < 0.0001).

In contrast to the atorvastatin treatment group, there was no significant association between lipid-lowering efficacy of simvastatin and CYP3A4*1G polymorphism (P > 0.05) (Table 2). After adjustment for age; gender; pretreatment TG, TC, LDL-C, and HDL-C levels; and genotype group, multiple regression analysis showed that CYP3A4*1G polymorphism was not associated with lipid-lowering efficacy of simvastatin.

Atorvastatin and simvastatin were well tolerated by all patients. No subject showed any elevation of aminotransferase or creatine phosphokinase after treatment, and no subjects reported skeletal muscle abnormalities or other notable safety concerns during the study.

Discussion

This is the first report on the frequency of CYP3A4*1G in a large-scale study of Chinese hyperlipidemic patients and the lipid-lowering efficacy of statins in individuals with different CYP3A4 genotypes. In this study, we found that the frequency of the CYP3A4*1G variant allele was 0.276 in 416 Chinese hyperlipidemic patients, which is the highest mutation of all the CYP3A4 alleles identified in a Chinese population. The frequency of mutant alleles and genotypes of this polymorphism was similar to the results reported previously in a Japanese population [12] and in a Chinese healthy population [13].

CYP3A4 is the most abundant isoenzyme of CYP450 in the human liver. It metabolizes numerous clinically, physiologically, and toxicologically important compounds. The expression of CYP3A4 varies 40-fold in individual human livers, and metabolism of substrates varies at least tenfold in vivo [17, 18]. Atorvastatin is administered in an active hydroxy-acid form, whereas simvastatin is administered as a prodrug requiring in vivo conversion to its active forms, including an active hydrolytic product, simvastatin hydroxy acid (SVA) [19, 20]. The metabolism of SVA in human liver microsomes is catalyzed primarily (≥80%) by CYP3A4/5 [21]. Atorvastatin is metabolized by CYP3A4 to ortho- and parahydroxylated derivatives [22]. Seventy percent of the HMG-CoA reductase inhibitory activity associated with atorvastatin has been attributed to its active metabolites, the ortho- and parahydroxylated derivatives, which are equipotent to the parent drug [23]. A significant increase in the plasma concentration of atorvastatin has also been reported after the concomitant use of itraconazole or erythromycin, both of which had marked inhibitory effects on CYP3A4 substrates [24, 25]. Therefore, it is plausible that variation in the CYP3A4 gene may lead to variation in CYP3A4 activity and, in turn, to the difference in the metabolism and, ultimately, the variable response of atorvastatin. Consistent with this postulation, we found in our study that subjects with the CYP3A4*1G/*1G genotype showed a higher total-cholesterol-lowering effect compared with those with CYP3A4*1/*1G genotype or CYP3A4*1/*1 genotype. Presumably, the CYP3A4*1G allele may decrease the activity of CYP3A4. The activity of CYP3A4 in mutant homozygotes (*1G/*1G) variant subjects was lower than in homozygous wild-type (*1/*1) subjects or heterozygous (*1/*1G) subjects, and decreased CYP3A4 activity would be predicted to cause slow drug metabolism and, ultimately, a higher plasma concentration of atorvastatin. Thus, our study showed that *1G expression seemed to increase the total-cholesterol-lowering effects of atorvastatin, probably resulting from a lower enzymatic activity of CYP3A4/*1G. Our finding is inconsistent with Hu et al.’s experimental observation [14]. They reported that subjects with the CYP3A4*18B/*18B genotype (actually called CYP3A4*1G/*1G in the study) had smaller mean Cmax of cyclosporine and faster mean clearance (CL) compared with subjects carrying the CYP3A4*1/*1 genotype. Their data indicated that the CYP3A4*1G mutant allele may be associated with the increased level of CYP3A4 activity. Therefore, further studies are required to confirm our findings.

In this study, the percentage reduction of LDL-C, TG, and HDL-C did not show a significant difference between the three genotypes. Because the subjects were recruited without knowing their genotypes, and after genotype assay, homozygous wild-type (*1/*1) subjects (107) were ten times more prevalent than mutant homozygotes (*1G/*1G) variant subjects (11). This affected sample power, which presumably explains the significant positive result only with TC but not LDL-C and TG.

For simvastatin, the result shows that CYP3A4*1G genetic polymorphism may have no significant effect on simvastatin treatment. The lack of association between lipid response to simvastatin and CYP3A4 genotypes may result, in part, from differences in the relative contributions of CYP3A4 protein in the metabolism of different statins. This study indicates that CYP3A4*1G genetic polymorphism may not necessarily be the major determinant in the lipid-lowering response to simvastatin.

In summary, CYP3A4*1G polymorphism is associated with the lowering of TC by atorvastatin in hyperlipidemic subjects. The study provides an important foundation and theoretical evidence for gene-directed rationalization and individualization of medication in hyperlipidemic patients. Further studies are required to evaluate the association of the CYP3A4*1G genotype with statin metabolism in healthy subjects.

Acknowledgements

This work was supported by grants from the Science Foundation for Distinguished Young Scholars of Henan Province (No. 074100510020) and the Engineering Project for Innovative Scholars of Henan Province (No. 2004020). The work was completed in the Henan key laboratory of molecular medicine. P.R. China. We thank Prof. Hu Dong-sheng (Department of Epidemiology, College of Public Health, Zhengzhou University, Zhengzhou, Henan, China) for helping us analyze data.

Conflict of interest statement

There is no conflict of interest.

Copyright information

© Springer-Verlag 2008