Simvastatin is one of the most widely used drugs for the treatment of hypercholesterolemia. Formulated as a prodrug, simvastatin lactone [simvastatin (SV)] undergoes metabolism to simvastatin hydroxyl acid [simvastatin acid (SVA)], the active form of the drug. SVA is a potent competitive inhibitor of the rate-limiting enzyme (HMG-CoA reductase) of intrahepatic cholesterol synthesis and thereby lowers plasma cholesterol concentrations. SV and SVA are primarily metabolized by CYP3A4/5 1, and increased plasma simvastatin concentrations have been associated with the use of concomitant medications that inhibit CYP3A metabolism 2,3. Likewise, polymorphisms that decrease CYP3A activity result in increased simvastatin concentration, exposure, and consequently pharmacodynamic effects. In a single-dose study of 20 mg SV, the 12-h area under the curve of the concentration-versus-time plot (AUC12) for homozygous carriers of the loss-of-function (LOF) CYP3A5*3 allele was 3.3-fold higher than for homozygous wild type (*1/*1) 4. Similarly, in a cohort of Whites receiving simvastatin, atorvastatin, or lovastatin (statins all primarily metabolized by CYP3A4/5) it was found that CYP3A5 nonexpressors (*3/*3) had greater reductions in total cholesterol compared with CYP3A5 expressors (*1/*1 or *1/*3) 5. The CYP3A4*22 decrease-of-function (DOF) allele has been associated with attenuated statin metabolism and with increased statin efficacy. CYP3A4 enzyme activity in livers with the CYP3A4*1/*1 wild-type genotype was 2.5-fold higher than in livers with the CYP3A4*22/*1 or CYP3A4*22/*22 genotypes 6. In addition, CYP3A4*22 carriers have been reported to require a 40% lower dose of simvastatin for optimal lipid control compared with wild type 6. Furthermore, CYP3A4*22 carriers from the Rotterdam study had significantly increased reductions in total and low-density lipoprotein-cholesterol with simvastatin treatment compared with noncarriers 7. Although relationships between CYP3A polymorphisms and clinical outcomes have been attributed to variation in simvastatin concentrations, the relationship between the CYP3A4*22 allele and in-vivo simvastatin concentrations has not yet been examined.
Currently, the FDA mandates drug labeling of simvastatin to include warnings recommending against the concomitant use of CYP3A4/5 inhibitors 8. Drug labels may evolve to include information regarding CYP3A genetic polymorphisms if sufficient evidence accrues establishing a clinically significant link with increased simvastatin exposure. The aim of this brief report is to determine the association between plasma simvastatin concentrations and the DOF CYP3A4*22 allele, alone and in combination with the LOF CYP3A5*3 allele.
Patients and methods
Investigators enrolled 1007 patients in the Cholesterol and Pharmacogenomics (CAP) study (clinicaltrials.gov identifier NCT00451828). The study’s primary aims were to assess the associations of DNA sequence variants with simvastatin pharmacodynamics 9. In total, 944 participants in the CAP study were genotyped, but complete data (CYP3A4/5 genotype and plasma simvastatin concentrations) was available for only 830 (555 self-reported Whites and 275 self-reported African-Americans) participants.
For White study participants, CYP3A5*3 (rs776746) status was determined using the Illumina HumanHap300 or HumanHap610-Quad (San Diego, California, USA) genotyping platforms. CYP3A5*3 status for African-American study participants was determined using TaqMan (Life Technologies, Grand Island, New York, USA) genotyping assay (C_26201809_30). CYP3A4*22 (rs35599367) allelic status was determined using TaqMan genotyping assay (C_59013445_10). To validate the rs776746 TaqMan assay, individuals of known genotype of each genotype class were included on each qPCR plate. These positive controls consisted of self-reported Whites for whom genome-wide genotype information was available. The rs35599367 TaqMan assay was validated in a similar manner using imputed genotype information. However, no minor allele homozygotes were available in our population to use as positive controls. No template controls were run on each plate as negative controls. A subset of samples (≤10%) were re-run, and no evidence of sample switching or of false genotyping was found.
Quantification of simvastatin and simvastatin acid
Plasma concentrations of SV and SVA were measured in CAP study participants after 6 weeks of 40 mg/day simvastatin. Compliance was monitored by pill count at 4 and 6 weeks after the start of statin treatment. As samples were collected ∼12 h after the final dose, only the compliance scores collected at the 6-week time-point were considered relevant for this analysis. Quantification was performed on a PE SCIEX API 365 (Framingham, Massachusetts, USA) triple–quadrupole mass spectrometer coupled to a liquid-chromatography column. Samples were evaluated on a Kromasil C18 (Saint Louis, Missouri, USA) column following the previously published protocol 10. Purified SV and SVA were gifted by Merck (New Jersey, USA), and serial dilutions were used to generate standard curves.
Statistical analyses were performed using Stata/IC-12 (StataCorp., College Station, Texas, USA) for Windows. All analyses were performed using nonparametric tests as data distributions were not normal.
On the basis of a-priori knowledge from in-vitro metabolism studies 4,6, we hypothesized the following: (i) simvastatin concentrations are higher in participants carrying at least one DOF CYP3A4*22 allele compared with wild type (*1/*1); (ii) simvastatin concentrations are higher for CYP3A5 nonexpressors (*3/*3) compared with CYP3A5 expressors (*1/*3 or *1/*1); and (iii) simvastatin concentrations are higher in combined CYP3A4/5 genotype-defined ‘poor metabolizers’ (PMs) compared with ‘intermediate metabolizers’ (IMs) and compared with ‘extensive metabolizers’ (EMs). Methodology regarding the combined CYP3A4/5 genotype-defined classification was first described by Elens et al.11 and has subsequently been utilized in the study of several CYP3A substrates 12,13.
All analyses were stratified by race. Potential confounding factors such as smoking status, sex, age, BMI, and compliance were tested for significant associations with simvastatin concentrations utilizing Mann–Whitney U-tests and Spearman’s rank correlation tests. In addition, associations with the genotype-defined groups were tested (Table 1) using Mann–Whitney U-tests and Kruskal–Wallis H-tests to compare medians of numeric variables between genotype-defined groups and Pearson’s χ2-tests to compare percentages between genotype-defined groups.
Median and interquartile ranges for SV and SVA were determined for genotype-defined groups (Figs 1 and 2). One-sided Mann–Whitney U-tests were performed for single-gene analyses comparing median values of SV and SVA between two genotype-defined groups. Directional hypotheses based on a-priori knowledge predicated the use of one-sided statistical tests. As BMI was associated with SV and SVA (P<0.05), both SV and SVA are lipophillic, and the average CAP participant was classified as overweight, all analyses were repeated with BMI-adjusted simvastatin concentrations 14.
As shown in Table 2, the allele frequencies of CYP3A4*22 and CYP3A5*3 were significantly higher in Whites, and distributions in the various genotype-defined groups were significantly different between Whites and African-Americans. In both races, CYP3A4*22 and CYP3A5*3 distributions were in Hardy–Weinberg equilibrium. The alleles were determined to be in partial linkage disequilibrium (LD), and the degree of LD was different for Whites and African-Americans (r2=0.06 and 0.14, respectively). As shown in Table 1, race, BMI, and smoking status had significant associations with genotype-defined groups. Race and BMI, however, were the only clinical characteristics with significant associations (P<0.05) with simvastatin concentrations, predicating that analyses be stratified by race and simvastatin concentrations be adjusted for BMI. Analyses were also repeated with concentrations adjusted for body weight.
Single-gene and combined-gene analyses of SV and SVA are presented in Figs 1 and 2. In the single-gene analyses of self-reported Whites, CYP3A5*3 allele status was not significantly associated with SV or SVA concentrations, and CYP3A4*22 carriers had 14% higher concentrations of SVA (P=0.04) and 20% higher concentrations of SV (P=0.06) compared with noncarriers. In the single-gene analyses of self-reported African-Americans, CYP3A4*22 carriers had 170% higher concentrations of SV than did noncarriers (P<0.01), but no significant difference was detected for concentrations of SVA; and CYP3A5 nonexpressors (CYP3A5*3/*3) had 33% higher concentrations of SV than CYP3A5 expressors (CYP3A5*1/*3 or *1/*1) (P=0.02), but no significant difference was detected for concentrations of SVA.
In the analysis of CYP3A4/5 genotype-defined groups (PMs, IMs, and EMs), SV concentrations appeared to decrease across the rank-ordered groups; however, no statistically significant associations were identified. Furthermore, similar trends were not observed for SVA concentrations. BMI adjustment had no appreciable effect in any analysis (results not shown).
Both White and African American CYP3A4*22 carriers had greater SVA than noncarriers (P<0.05); however, this effect was much more remarkable in the African-Americans than Whites (170 vs. 14% increase, respectively). A nonstatistically significant trend of greater SV in CYP3A4*22 carriers was also observed. The effect of CYP3A5 status was significant only for African-Americans; SV was 33% greater in CYP3A5 nonexpressors (*3/*3) compared with CYP3A5 expressors (*1/*3 or *1/*1) (P=0.02).
This report describes associations between in-vivo simvastatin concentrations and CYP3A5*3 and CYP3A4*22 alleles, and the observed associations were consistent with the known functional effects of these alleles 4,6. The DOF CYP3A4*22 allele was associated with increased simvastatin concentrations for both self-reported Whites and self-reported African-Americans. However, CYP3A5 was influential in only the self-reported African-Americans. Although our results are not unexpected, to date this is the first report that has specifically tested the associations between plasma simvastatin concentration and CYP3A4*22, alone or in the combined CYP3A4/5 genotype-defined classification.
Observed differences in CYP3A allele frequencies and LD between African-Americans and Whites are consistent with previously reported allele frequency analyses 15. We observed that the majority (nearly 90%) of African-Americans were categorized as EMs in the combined CYP3A4/5 genotype-defined classification system (Fig. 2). We previously reported that simvastatin treatment has reduced efficacy for low-density lipoprotein-cholesterol lowering in African-Americans compared with Whites 9. This difference may partially be attributed to the disproportionate number of EMs in African-Americans; however, further investigation is warranted.
Our analyses failed to identify a statistically significant relationship between CYP3A5 genotype and plasma SV or SVA concentrations in Whites. This is consistent with the analyses reported by Kim et al. 4 in which a statistically significant difference was determined for simvastatin concentration AUC12 but not for 12-h postdose concentration. The LOF CYP3A5*3 allele is the major allele in Whites, and it is likely that variation in other similar enzymes, such as CYP3A4 or CYP3A7, may compensate for the loss of CYP3A5 function associated with the *3 allele.
Of the covariates tested, we only found evidence of a modest association between BMI and statin concentrations. This finding was not surprising because SV and SVA are lipophilic 16,17, the average BMI in the CAP study classifies as overweight (i.e. average BMI was 27.8), and BMI can alter the pharmacokinetic parameters of lipophilic substances. However, as the BMI range of the cohort was quite narrow and the effect size of BMI on SV and SVA was marginal, it was not unexpected that BMI adjustment did not impact the observed genotype–concentration associations. Similarly, all analyses were repeated with adjustments for body weight and the observed genotype–concentration associations were not significantly altered (results not shown).
There are several limitations to the present report. It is based on a retrospective pharmacokinetic analysis in a study designed primarily for measuring pharmacodynamic endpoints and did not examine polymorphisms in other genes known to affect simvastatin pharmacokinetics (e.g. SLC01B1). In addition, racial categories were self-reported, and although race stratification does account for racial differences in LD, it does not account for unbalanced allele status among polymorphism-carrier groups. Finally, some genotype-determined groups were inadequately represented (e.g. only two of 275 African-American participants represented the CYP3A4/5 PM group).
Nevertheless, the study has several strengths for investigating influence of CYP3A: the overall sample size was large, the simplicity of the genotype groupings allowed for utilization of robust nonparametric statistical analyses, the sex profile was well balanced, the ranges of BMI and age were marginal, and simvastatin dose, strength, and duration were homogenous. Furthermore, the study design prohibited the use of concomitant medications and certain foods that could induce or inhibit CYP3A.
The genotype–concentration associations reported here are preliminary observations and are not intended to guide the dose selection of simvastatin. However, these results may be combined with additional data to account for ancestry, LD, and other polymorphic genes involved in the pharmacokinetics of simvastatin, possibly providing a more predictive model in the future.
Here, we presented the first report specifically testing the associations between plasma simvastatin concentration and CYP3A4*22, alone or in the combined CYP3A4/5 genotype-defined classification. Our analyses revealed associations between simvastatin concentrations and CYP3A4*22 in both racial categories and an association for CYP3A5 in only the self-reported African-Americans.
Guidance for statin pharmacotherapy is of increasing relevance given the dramatic expansion of statin-use recommendations outlined in the clinical guidelines recently issued by the American Heart Association 18. Although CYP3A genotype is currently not considered when prescribing statins, future guidance and recommendations may be derived in part from models incorporating genotype status along with concomitant use of enzyme inhibitors or inducers.
The authors thank Elizabeth Theusch, Sarah King, Larry Wong, and Ann Wang for their assistance in genotyping and data generation.
This research was supported by the following grants from the National Institutes of Health: U19 HL069757, R01 HL104133, K23 GM100372, and U01 GM092655.
Conflicts of interest
There are no conflicts of interest.
1. Prueksaritanont T, Ma B, Yu N. The human hepatic metabolism
hydroxy acid is mediated primarily by CYP3A, and not CYP2D6. Br J Clin Pharmacol 2003; 56:120–124.
2. Christians U, Jacobsen W, Floren LC. Metabolism
and drug interactions of 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors in transplant patients: are the statins mechanistically similar? Pharmacol Ther 1998; 80:1–34.
3. Corsini A, Bellosta S. Drug–drug interaction with statins. Expert Rev Clin Pharmacol 2008; 1:105–113.
4. Kim KA, Park PW, Lee OJ, Kang DK, Park JY. Effect of polymorphic CYP3A5
genotype on the single-dose simvastatin
pharmacokinetics in healthy subjects. J Clin Pharmacol 2007; 47:87–93.
5. Kivisto KT, Niemi M, Schaeffeler E, Pitkälä K, Tilvis R, Fromm MF, et al.. Lipid-lowering response to statins is affected by CYP3A5
6. Wang D, Guo Y, Wrighton SA, Cooke GE, Sadee W. Intronic polymorphism in CYP3A4
affects hepatic expression and response to statin drugs. Pharmacogenomics J 2011; 11:274–286.
7. Elens L, Becker ML, Haufroid V, Hofman A, Visser LE, Uitterlinden AG, et al.. Novel CYP3A4
intron 6 single nucleotide polymorphism is associated with simvastatin
-mediated cholesterol reduction in the Rotterdam study. Pharmacogenet Genomics 2011; 21:861–866.
8. Merck Sharp & Dohme Corp. Simvastatin
prescribing information; 1999–2014. Available at: http://www.merck.com/product/usa/pi_circulars/z/zocor/zocor_pi.pdf
. [Accessed 14 April 2014].
9. Simon JA, Lin F, Hulley SB, Blanche PJ, Waters D, Shiboski S, et al.. Phenotypic predictors of response to simvastatin
therapy among African-Americans and Caucasians: the Cholesterol and Pharmacogenetics
(CAP) Study. Am J Cardiol 2006; 97:843–850.
10. Zhao JJ, Xie IH, Yang AY, Roadcap BA, Rogers JD. Quantitation of simvastatin
and its beta-hydroxy acid in human plasma by liquid-liquid cartridge extraction and liquid chromatography/tandem mass spectrometry. J Mass Spectrom 2000; 35:1133–1143.
11. Elens L, Bouamar R, Hessalink DA, Haufroid V, van der Heiden IP, van Gelder T, van Schaik RH. A new functional CYP3A4
intron polymorphism significantly affects tacrolimus pharmacokinetics in kidney transplant recipients. Clin Chem 2011; 57:1574–1583.
12. Kitzmiller JP, Sullivan DM, Phelps MA, Wang D, Sadee W. CYP3A4
/5 combined genotype analysis for predicting statin dose requirement for optimal lipid control. Drug Metabol Drug Interact 2013; 28:59–63.
13. Kurzawski M, Dabrowska J, Dziewanowski K, Domański L, Perużyńska M, Droździk M. CYP3A5
, but not ABCB1 polymorphisms affect tacrolimus dose-adjusted trough concentrations in kidney transplant recipients. Pharmacogenomics 2014; 15:179–188.
14. Blouin RA, Warren GW. Pharmacokinetic considerations in obesity. J Pharm Sci 1999; 88:1–7.
15. Wang D, Sadee W. The making of a CYP3A biomarker panel for guiding drug therapy. J Pers Med 2012; 2:175–191.
16. Davidson MH, Robinson JG. Lipid-lowering effects of statins: a comparative review. Expert Opin Pharmacother 2006; 7:1701–1714.
17. Ballantyne CM. Clinical lipidology: a companion to Braunwald’sheart disease
2009:1st ed..Philadelphia, PA:Saunders/Elsevier;584.
18. Stone NJ, Robinson J, Lichtenstein AH, Bairey Merz CN, Blum CB, Eckel RH, et al.. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2014; 63Pt B2889–2934.