Intrinsic Factors Influencing Simvastatin and Simvastatin Acid Pharmacokinetics: Age‐Related Studies in Thai Adults and Cross‐Population Comparisons

ABSTRACT

Intrinsic factors such as age and ethnicity influence statin pharmacokinetics (PK) and may increase the risk of muscle‐related adverse events. This study investigated the effect of age on the PK of simvastatin (SV) and its active metabolite, simvastatin acid (SVA), in healthy Thai subjects using an LC–MS/MS assay while accounting for key SLCO1B1 variants known to affect statin exposure. We further compared our results with published data across multiple ethnic populations via meta‐analysis and conducted an exploratory pharmacogenetic analysis of SVA disposition‐related genes to identify potential candidate variants contributing to PK differences observed between Thai and Caucasian populations. The pharmacokinetic results showed that aging significantly increased SV C_max but did not alter SVA exposure, while SLCO1B1 decreased and poor functions were associated with higher SVA exposure. The meta‐analysis revealed that Thai subjects showed significantly higher SVA levels than Caucasians, Chinese, and Japanese. The pharmacogenetic analysis identified functional or promoter variants in SLCO1B1 and PON genes with allele frequency differences directionally consistent with the observed PK differences between Thais and Caucasians. Our results may partly explain the higher incidence of SV‐associated rhabdomyolysis previously reported in Thais compared with Caucasians and highlight the relevance of these intrinsic factors in SV and SVA PK, with implications for dosing considerations in Thai patients. Additional studies are warranted to further support the involvement of these polymorphic variants in the observed interethnic differences in SVA PK and adverse outcomes.

Study Highlights

• What is the current knowledge on the topic?

  • ○
    Age and ethnicity can influence statin pharmacokinetics and may increase the risk of statin‐related myotoxicity. A disproportionately high number of simvastatin‐associated rhabdomyolysis cases have been reported in Thai patients, many of advanced age, compared with Western cohorts. Polymorphisms in statin disposition‐related genes, particularly SLCO1B1, are known contributing factors to interindividual and interethnic variability in statin exposure and associated risks.
• What question did this study address?

  • ○
    How does aging affect the pharmacokinetics of simvastatin and its active metabolite, simvastatin acid, in healthy Thai adults?
  • ○
    Are there ethnic differences in simvastatin and simvastatin acid exposures between Thais and other populations, particularly Caucasians, which could partially explain the reported higher incidence of simvastatin‐associated adverse events?
  • ○
    What potential genetic variants may contribute to the observed interethnic differences?
• What does this study add to our knowledge?

  • ○
    Aging increased the exposure of simvastatin but not simvastatin acid, unlike SLCO1B1 polymorphisms, which altered the exposure of simvastatin acid but not simvastatin.
  • ○
    Plasma exposures of simvastatin acid in Thais were significantly higher than those reported in Caucasians, a difference aligned with the higher incidence of simvastatin‐associated rhabdomyolysis reported in Thais.
  • ○
    Polymorphisms in SLCO1B1 and PON1/2/3 genes with marked allele frequency differences between Thais and Caucasians may partially explain the observed interethnic differences.
• How might this change clinical pharmacology or translational science?

  • ○
    This study highlights age and ethnicity as important intrinsic factors for simvastatin and simvastatin acid pharmacokinetics, with implications for guiding optimal dosing of simvastatin in Thai patients. The findings also provide a foundation for future studies to elucidate the roles of SLCO1B1 and PON1/2/3 variants in statin disposition.

1. Introduction

Statins, or 3‐hydroxy‐3‐methylglutaryl coenzyme A (HMG‐CoA) reductase inhibitors, are the first‐line pharmacologic treatment for hypercholesterolemia, a major risk factor for atherosclerotic cardiovascular disease, and its associated morbidity and mortality [1]. Simvastatin (SV) is one of the most widely prescribed statins worldwide, offering cost‐effective lipid‐lowering therapy [1]. Unlike many other statins, SV is a lactone prodrug that requires biotransformation by carboxylesterase (CES) and paraoxonase (PON)‐mediated hydrolysis to form its active open‐acid metabolite, simvastatin acid (SVA) [2]. Both SV and SVA undergo extensive metabolism by cytochrome P450 (CYP) 3A4/5 [3, 4]. SVA is additionally metabolized by UDP‐glucuronosyltransferase (UGT) 1A1/3, followed by spontaneous lactonization back to SV [5]. Importantly, SVA, but not SV, is a substrate of the hepatic uptake transporter organic anion transporting polypeptide (OATP) 1B1/3 encoded by the SLCO1B1/3 genes [6, 7]. Polymorphisms in SLCO1B1 can significantly influence the pharmacokinetics (PK) of SVA [8, 9] and other open acid statins such as atorvastatin and rosuvastatin [10].

In Thailand, SV is the most frequently prescribed statin [11]. However, SV and SVA PK data in the Thai population remain scarce, with only two publications: a full PK bioequivalence study in healthy young subjects and a single‐time‐point plasma PK study in patients with dyslipidemia or coronary artery disease [9, 12]. A related PK study in healthy Thai subjects using a microdose of atorvastatin and midazolam demonstrated increased oral exposure in elderly participants [13], suggesting reduced CYP3A activity with aging. In the same study, the authors reported no significant effects on the oral PK profiles of rosuvastatin and pitavastatin, both of which are not CYP3A substrates [13]. While increased SV exposure may therefore be anticipated in older adults, the effect of age on SVA remains uncertain, given its complex dependence on both formation and elimination pathways. Considering that a large portion of patients receiving statins are elderly [9, 11, 14, 15], knowledge of age‐related changes in statin PK is of particular importance. Although the effect of age was reported in hypercholesterolemic patients following SV administration, this study used a bioassay that measured HMG‐CoA reductase inhibitory activity [16]. To date, no study has directly examined the impact of age on SV and SVA PK using a specific quantitative assay.

In addition, Thai national pharmacovigilance data (Thai Vigibase) have reported a disproportionally high number of rhabdomyolysis cases among patients receiving lower SV doses than Western cohorts [14]. Yet, no comprehensive PK comparison between Thais and Caucasians has been conducted. Although racial/ethnic differences in statin PK have been documented, particularly for atorvastatin, rosuvastatin, and, to a lesser extent, SVA, with higher exposure in Chinese and Japanese than in Caucasians [17], it remains unclear whether such findings extend to Thais, as systematic cross‐ethnic comparisons are lacking. For rosuvastatin, such differences led to a lower recommended starting dosage in Asians [18]. Pharmacogenetic analyses have attributed part of the ethnic variability to a higher frequency of the ABCG2 c.421C>A variants in Asians, which reduces breast cancer resistance protein (BCRP) activity, thereby increasing exposure to its substrates [17]. However, unlike atorvastatin and rosuvastatin, SVA is not a BCRP substrate [19], suggesting that additional genetic or other intrinsic factors may underlie interethnic variability.

In this study, we aimed to (1) evaluate the effect of age on the PK of SV and SVA in healthy Thai subjects, using a highly specific assay, while accounting for key SLCO1B1 variants recognized by the Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline [20]; (2) compare the PK of SV and SVA in Thais with data from other ethnic populations, particularly Caucasians, to help explain the reported higher incidence of SV‐associated adverse events; and (3) perform an exploratory pharmacogenetic analysis of genes relevant to SVA disposition using publicly available genome databases to identify candidate variants potentially contributing to interethnic PK differences.

2. Methods

2.1. PK Study in Young Adults and Elderly

2.1.1. Participants and Study Design

This study was reviewed and approved by the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand (IRB No. 156/65). The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines. Written informed consent was obtained from all participants before study initiation. The study was conducted at the Maha Chakri Sirindhorn Clinical Research Center, Faculty of Medicine, Chulalongkorn University.

Fifty healthy participants were enrolled and stratified into two groups: healthy young adults, aged 20–40 years, and healthy elderly individuals, aged 60–80 years. The number of participants per group was estimated to provide 80% power at a two‐sided alpha level of 5% to detect a 50% difference in SV AUC between two studied groups, based on reported SV PK variability [12]. A full list of the inclusion and exclusion criteria is provided in Table S1.

All participants received a single oral dose of 20 mg SV. Venous blood samples were collected into EDTA tubes to minimize ex vivo SV hydrolysis by PON at various time points. Following blood collection, the EDTA tubes were placed on ice in a cool box, and plasma was separated by centrifugation within 15 min of collection. The plasma samples were then transferred into microcentrifuge tubes containing a buffer solution of 1 M ammonium acetate (pH 4.5) at a ratio of 100:5 (plasma:buffer), immediately mixed to minimize potential interconversion between SV and SVA, and subsequently stored at −80°C until analysis. The buffy coat fraction from each sample was retained for genotyping and stored at −20°C until analysis.

Safety assessments were performed by monitoring participants for signs and symptoms of myalgia, rhabdomyolysis, hepatitis, abdominal pain, and nausea during the PK study period, with follow‐up for 7–10 days after the study completion.

2.1.2. Bioanalysis

Plasma concentrations of SV and SVA were determined using ultra‐high performance liquid chromatography–tandem mass spectrometry (UHPLC–MS/MS). This method was adapted from a previous study [21] and conformed to the regulatory bioanalytical method validation requirements [22], as described earlier [9]. The lower limit of quantitation was 0.05 ng/mL for SV and 0.20 ng/mL for SVA.

2.1.3. PK Analysis

PK parameters were analyzed using non‐compartmental analysis. The maximum plasma concentration (C_max) and time to maximum plasma concentration (t _max) were directly obtained from the individual plasma drug concentration–time curve. The area under the plasma concentration–time curve from time zero to the last measurable time point (AUC_0–t) was calculated using the log‐linear trapezoidal method. The terminal elimination rate constant (k _el) was estimated by log‐linear least‐squares regression of the terminal phase of the concentration–time curve and used to extrapolate total exposure (AUC_0–inf). The terminal half‐life (t _1/2) was determined as ln (2)/k _el.

2.1.4. DNA Preparation and Genotyping

Genomic DNA was extracted from the buffy coat samples using the QIAamp Blood Mini Kit (Qiagen, Hilden, Germany). DNA quantity and quality were assessed using the NanoDrop One Microvolume UV–Vis Spectrophotometer (Thermo Scientific). Genotyping SLCO1B1 single nucleotide polymorphisms (SNPs) was performed using the MassARRAY System (Agena Bioscience). Target SNPs were selected based on known associations with statin PK according to the CPIC guidelines [20] (Figure S1).

2.1.5. Statistical Analysis

Statistical analysis was performed using SPSS version 29.0 (IBM Corp., Armonk, NY, USA), and graphs were generated using GraphPad Prism version 10.4 (GraphPad Software, San Diego, CA, USA). Continuous variables were summarized as median (interquartile range, IQR), and categorical variables as counts and percentages. The Mann–Whitney U test was used for continuous variables, and Fisher's exact test for categorical variables.

For PK parameters (AUC, C_max, and t _1/2), results were expressed as geometric mean with a 95% confidence interval (CI). Values for t _max were reported as the median (range). Geometric mean ratio (GMR) with corresponding 90% CI comparing healthy elderly and young adults was calculated by log‐transforming the mean, determining the arithmetic difference and its 90% CI, and then back‐transforming to the original scale.

Stepwise multivariable linear regression was used to evaluate the contribution of SLCO1B1 phenotype and non‐genetic factors (age, body weight, gender, and serum chemistry) to key PK parameters (C_max and AUC) of SV and SVA, as well as the metabolite‐to‐parent ratios. A subgroup analysis restricted to participants with normal‐function SLCO1B1 phenotypes was conducted to minimize confounding from SLCO1B1 polymorphisms.

2.2. Meta‐Analysis of PK Studies in Multiple Ethnicities

2.2.1. Data Source and Collection

Published PK studies of SV were identified through a PubMed search using the terms “simvastatin AND pharmacokinetic” covering January 2000 to December 2022. A total of 878 articles were screened for eligible full PK studies in healthy participants based on the criteria presented in Figure 1. These criteria were applied to minimize data heterogeneity across studies and enhance the reliability of cross‐ethnic comparisons. Since most eligible studies reported arithmetic means, these values were used in the meta‐analyses to maintain consistency across datasets.

FIGURE 1.

Flow diagram summarizing the screening process and selection criteria for inclusion of studies and pharmacokinetic (PK) dataset in the meta‐analyses.

2.2.2. Data Analysis

For each eligible study, the C_max and AUC ratios of SVA to SV were extracted and stratified by ethnicity for cross‐ethnic comparisons of SVA disposition. Meta‐analyses were performed using the ratio of means (RoM) approach [23]. The RoM and corresponding standard error (SE) were calculated, log‐transformed, and pooled via inverse variance (IV) method. Subgroup results were reported as pooled estimates with 95% CIs. DerSimonian and Laird random‐effects model was applied to account for both within‐ and between‐study variability. Statistical heterogeneity among studies within each ethnic group was assessed using the I ² and Chi‐square tests. Subgroup and pairwise ethnic differences were assessed using the Chi‐square test. All analyses were conducted using Cochrane Review Manager (RevMan), version 5.4.

2.3. Population‐Level Variants in SV PK‐Related Genes

2.3.1. Allele Frequency Analysis

Allele frequencies (AFs) of SLCO1B1 and PON1/2/3 variants implicated in the formation and elimination of SVA [2, 6, 7], were analyzed between Thai and non‐Thai populations. Thai AFs were obtained from the Thai Genome Reference Database [24]. Non‐Thai AFs were obtained from the Genome Aggregation Database (gnomAD) v3.1.2 [25]. The non‐Finnish European (NFE) subset was used to represent the Caucasian cohorts in the studies included in our meta‐analyses (see Results). Specific AF data for the Chinese and Japanese populations were unavailable in gnomAD; therefore, AF difference testing was not conducted for these groups.

Differences in AFs between Thais and Caucasians were tested using Fisher's exact test. The significance level was set at p value < 0.01 after correction for multiple testing using the Benjamini–Hochberg procedure with a false discovery rate threshold of 0.05.

2.3.2. Expression Quantitative Trait Loci Analysis

The functional consequences of genetic variants on gene expression were evaluated using expression quantitative trait loci (eQTL) data from GTEx v8 [26]. Only eQTL effects in liver and small intestine, the key tissues involved in SV disposition, were considered.

A significant eQTL effect was defined as p value < 0.05, with a posterior probability (m value) > 0.9 for the presence of an effect in the relevant tissue [27]. The normalized effect size (NES) reported in GTEx was used to explain the magnitude and direction of the regulatory effect of each variant on gene expression.

3. Results

3.1. PK Study in Young Adults and Elderly

Fifty healthy participants were enrolled in this study, and the baseline characteristics by age group are shown in Table S2. The median ages for the young adult and elderly groups were 30 years (range 23–40 years) and 65 years (range 60–77 years), respectively. All participants were considered healthy, based on their serum biochemistry, hematology, and urinalysis results, although the elderly group showed higher total cholesterol and LDL‐C values compared to young adults. The genotype distribution of SLCO1B1 (CPIC guideline [20]) was similar between the two groups, with few participants in the young and elderly groups classified as decreased and poor function phenotypes (Table 1). Moreover, a single oral dose of 20 mg SV was well tolerated by all participants, and no adverse events or signs of adverse effects were observed.

TABLE 1.

Distribution of SLCO1B1 genotype and phenotype among the study participants.

Genotype/phenotype	Number of participants
	Young adults, n = 25	Elderly, n = 25	All participants a , n = 50
SLCO1B1 genotype b [SNP rsID (Ref>Alt allele)]
rs11045819 (c.463C>A)	25/0/0	25/0/0	50/0/0 (0)
rs183501729 (c.757C>T)	25/0/0	25/0/0	50/0/0 (0)
rs2306283 (c.388A>G)	0/8/17	1/7/17	1/15/34 (0.83)
rs34671512 (c.1929A>C)	22/3/0	25/0/0	47/3/0 (0.03)
rs373327528 (c.211G>A)	25/0/0	25/0/0	50/0/0 (0)
rs4149056 (c.521T>C)	20/4/1	22/3/0	42/7/1 (0.09)
rs56061388 (c.245T>C)	25/0/0	25/0/0	50/0/0 (0)
rs56199088 (c.1964A>G)	25/0/0	25/0/0	50/0/0 (0)
rs71581941 (c.1738C>T)	25/0/0	25/0/0	50/0/0 (0)
rs72559745 (c.467A>G)	25/0/0	25/0/0	50/0/0 (0)
rs72559747 (c.1007C>G)	25/0/0	25/0/0	50/0/0 (0)
Predicted SLCO1B1 phenotype c
Normal function (*1/*1, *1/*20, *1/*37, *20/*37, *37/*37)	20	22	42
Decreased and poor functions (*15/*37, *15/*15)	5	3	8

Abbreviations: Alt, alternative allele; Ref, reference allele.

^a Numbers in parentheses are alternative allele frequencies.

^b Numbers of participants are represented as wild type/heterozygous/homozygous variants.

^c Assignment of phenotype is following CPIC guideline for SLCO1B1 [20]—see Figure S1: No participants in this study carry *5 and *14 haplotypes.

Following the oral administration of SV, the plasma concentrations of SV and SVA declined rapidly in both young and elderly participants, with 12‐h post‐dose levels detectable in nearly all subjects, but the 24‐h post‐dose levels were beyond the detection limit in several participants (Figure 2A,B). The t _1/2 values for SV (~4–5 h) were comparable to or slightly longer than SVA (~3–4 h) (Table 2). The AUC_0–12h values accounted for the majority (~80%) of AUC_0–inf in both groups. Across all participants, SVA exhibited higher exposure than SV. The SVA/SV C_max and AUC ratios were approximately 1.1–1.6 and 2–3, respectively (Table 2).

FIGURE 2.

Plasma concentration–time profiles of simvastatin (SV, panel A) and simvastatin acid (SVA, panel B) following the oral administration of a single dose of 20 mg SV to healthy Thai subjects. (●) Young adults, (■) Elderly individuals. Bars represent the standard deviation (SD), n = 25 per group, except at 24‐h post‐dose, where n = 24 (young) and n = 19 (elderly) for SV, and n = 15 (young) and n = 12 (elderly) for SVA due to assay sensitivity limitations. The inset in panel A shows scatter plots of SV C_max values from all 50 subjects. *Statistically significant difference between groups, p < 0.05.

TABLE 2.

PK parameters of SV and SVA in all participants and in the subgroup of SLCO1B1 normal function subjects following a single 20 mg oral dose of SV.

PK parameter	All participants			Participants with SLCO1B1 normal function
	Geometric mean (95% CI)		GMR (elderly/young adult) (90% CI)	Geometric mean (95% CI)		GMR (elderly/young adult) (90% CI)
	Young adults (n = 25)	Elderly (n = 25)		Young adults (n = 20)	Elderly (n = 22)
SV
t max (h) a	1.5 (0.5–6.0)	1.5 (0.5–4.0)	N/A	1.5 (0.5–6.0)	1.5 (0.5–4.0)	N/A
Cmax (ng/mL)	3.10 (2.45–3.92)	4.73 (3.90–5.74)*	1.53 (1.19–1.96)	3.12 (2.33–4.18)	4.65 (3.80–5.68)*	1.49 (1.12–1.97)
AUC0–12h (ng·h/mL)	14.33 (11.42–17.98)	16.51 (13.52–20.16)	1.15 (0.90–1.47)	14.02 (10.54–18.66)	15.79 (12.97–19.22)	1.13 (0.85–1.48)
AUC0–inf (ng·h/mL) b	18.27 (13.92–23.98)	19.25 (15.65–23.67)	1.05 (0.80–1.39)	17.81 (12.78–24.84)	18.51 (15.07–22.74)	1.04 (0.76–1.41)
t 1/2 (h) b	4.76 (3.93–5.78)	4.02 (3.29–4.91)	N/A	4.70 (3.79–5.82)	4.20 (3.37–5.25)	N/A
SVA
t max (h) a	4.0 (2.0–8.0)	3.0 (2.0–6.0)	N/A	4.0 (2.0–8.0)	3.5 (2.0–6.0)	N/A
Cmax (ng/mL)	5.02 (3.78–6.65)	5.14 (4.03–6.56)	1.02 (0.76–1.39)	4.35 (3.32–5.68)	5.07 (4.15–6.21)	1.17 (0.89–1.53)
AUC0–12h (ng·h/mL) c	32.60 (24.54–43.32)	34.06 (26.61–43.60)	1.04 (0.77–1.42)	28.26 (21.56–37.04)	33.29 (27.92–39.69)	1.18 (0.91–1.52)
AUC0–inf (ng·h/mL) d	44.91 (31.76–63.52)	41.03 (32.04–52.54)	0.91 (0.65–1.28)	38.73 (26.70–56.18)	40.20 (33.41–48.36)	1.04 (0.77–1.41)
t 1/2 (h) d	3.65 (2.92–4.57)	3.40 (2.73–4.24)	N/A	3.58 (2.74–4.67)	3.51 (2.64–4.66)	N/A
SVA/SV ratio
Cmax	1.62 (1.19–2.21)	1.09 (0.87–1.36)*	0.67 (0.49–0.92)	1.39 (1.07–1.82)	1.09 (0.87–1.37)	0.78 (0.59–1.04)
AUC0–12h c	2.18 (1.67–2.85)	2.06 (1.67–2.54)	0.95 (0.72–1.25)	1.91 (1.50–2.44)	2.11 (1.74–2.56)	1.10 (0.86–1.42)
AUC0–inf e	2.68 (2.13–3.36)	2.14 (1.71–2.69)	0.80 (0.62–1.04)	2.34 (2.00–2.73)	2.19 (1.76–2.71)	0.94 (0.75–1.17)

Abbreviation: N/A, not applicable.

^*Statistically significant differences compared with young adults (p < 0.05).

^a Data are presented as medians (range).

^b n = 22 in the young adult group, 23 in the elderly group, for analysis with all participants; 18 in the young adult group, 20 in the elderly group, for subgroup analysis.

^c n = 24 in both the young adult and elderly group, for analysis with all participants; 19 in the young adult group, 21 in the elderly group, for subgroup analysis.

^d n = 20 in the young adult group, 24 in the elderly group, for analysis with all participants; 15 in the young adult group, 21 in the elderly group, for subgroup analysis.

^e n = 19 in the young adult group, 22 in the elderly group, for analysis with all participants; 15 in the young adult group, 19 in the elderly group, for subgroup analysis.

When comparing age groups, elderly participants had significantly higher SV C_max (~1.5‐fold, p < 0.05) than young adults (Figure 2A, Table 2). In contrast, SVA C_max did not differ between age groups (Figure 2B, Table 2). Consequently, the SVA/SV C_max ratio was significantly lower in the elderly (0.67‐fold vs. young adults) (Table 2). The AUC values of SV and SVA, as well as the SVA/SV AUC ratio, did not differ between age groups (Table 2).

Multivariate linear regression analysis identified age and SLCO1B1 function as contributors to variability in PK of SV and SVA, respectively. Increasing age was positively associated with elevated SV C_max (p < 0.05), whereas reduced or poor SLCO1B1 function was associated with increased SVA C_max and AUC, as well as higher SVA/SV C_max and AUC ratios (p < 0.05) (Table S3). Gender also appeared to influence SVA C_max, although no significant associations were observed for SV nor SVA/SV C_max or AUC. In subgroup analyses restricted to participants with normal‐function SLCO1B1 phenotypes, elderly participants continued to show a 1.5‐fold higher SV C_max compared with young adults (Table 2). However, the SVA/SV C_max ratio did not differ significantly between the two groups, although a decreasing trend was observed in the elderly (Table 2).

3.2. Meta‐Analysis of PK in Multiple Ethnicities

Of the 17 PK datasets gathered from the literature and selected for the meta‐analysis, ten datasets were in Caucasians (n = 250), primarily non‐Finnish Europeans; five were in Chinese (n = 129); one was in Japanese (n = 40); and one was in Thais (n = 18) (Table 3). The participants in these studies received 20–60 mg of oral SV and were healthy subjects aged 17–59 years (Table 3). Except for the study in Thais, studies in all other ethnicities showed higher SV C_max and/or AUC values than SVA over the studied dose range (Figure S2A,B). Both the published [12] and current Thai studies (total n = 68) showed consistently higher C_max and AUC values for SVA than for SV (Figure S2A,B). Across these studies, there was moderate to high interindividual variability in the AUC_0–inf of SV and SVA. To enable a better comparison across different dose levels, the C_max and AUC values for SVA were normalized by the respective SV values and used for the subsequent meta‐analysis.

TABLE 3.

Summary of key participant characteristics and SV and SVA PK parameters in prior studies.

Author, year	Country	Ethnicity	Sample size	Age a	Male (%)	Dose (mg)	SD/MD	SV		SVA
								Cmax (ng/mL)	AUC0–inf (ng·h/mL)	Cmax (ng/mL)	AUC0–inf (ng·h/mL)
PK dataset included for meta‐analyses								Arithmetic mean ± standard deviation (% coefficient of variation)
Lohitnavy et al., 2004 [12]	Thailand	Thais	18	20.4 ± 0.8	100	40	SD	7.75 ± 3.42 (44%)	37.45 ± 17.9 (48%)	20.18 ± 17.59 (87%)	132.16 ± 113.5 (86%)
Becquemont et al., 2007 [28]	France	Caucasians	12	19–43	58	40	SD	11.6 ± 6.3 (54%)	40.6 ± 13.6 b (33%)	2.4 ± 1.3 (54%)	23.1 ± 12.9 b (56%)
Dawra et al., 2019 [29]	Belgium	Caucasians (83%) Blacks (11%) Other (6%)	18	31.3 ± 9.0	56	40	SD	9.283 ± 5.67 (61%)	43.78 ± 19.23 (44%)	2.605 ± 2.62 (101%)	31.44 ± 23.13 (74%)
Groenendaal‐van de Meent et al., 2022 (dataset 1) [30]	Germany	Caucasians (96.4%) Other (3.6%)	28	40.1 ± 11.5	57.1	40	SD	6.185 ± 4.891 (79%)	38.703 ± 23.827 (62%)	2.168 ± 1.500 (69%)	31.628 ± 17.858 (56%)
Groenendaal‐van de Meent et al., 2022 (dataset 2) [30]	Germany	Caucasians (95.8%) Other (4.2%)	24	45.0 ± 9.9	50.0	40	SD	6.52 ± 4.45 (68%)	37.74 ± 26.11 (69%)	2.12 ± 1.31 (62%)	24.31 ± 13.36 (55%)
Hasunuma et al., 2016 [31]	USA	Caucasians	40	25.7 ± 4.0	100	20	SD	4.26 ± 2.76 (65%)	16.9 ± 11.0 (65%)	1.09 ± 0.51 (47%)	13.2 ± 9.8 (74%)
Macha et al., 2014 [32]	Germany	Caucasians (94%) Blacks (6%)	18	20–50	67	40	SD	9.9 ± 6.7 (68%)	40.4 ± 23.1 (57%)	1.9 ± 1.3 (68%)	21.7 ± 15.6 (72%)
Poller et al., 2019 [33]	United Kingdom	Caucasians	20	30.5 ± 7.67	100	20	SD	4.60 ± 3.02 (66%)	17.5 ± 13.4 (77%)	0.678 ± 0.340 (50%)	7.50 ± 4.25 (57%)
Ucar et al., 2004 [34]	Sweden	Caucasians	12	17–31	100	60	SD	18.7 ± 16.3 (87%)	88.8 ± 51.6 (58%)	3.5 ± 1.7 (49%)	33.5 ± 12.5 (37%)
Winsemius et al., 2014 [35]	Germany	Caucasians (98%) Asians (2%)	85	21–55	60	40	SD	9.2 ± 5.8 (63%)	28.7 ± 13.0 (45%)	2.2 ± 1.6 (73%)	22.4 ± 18.1 (81%)
							MD	9.9 ± 7.0 (71%)	30.0 ± 12.0 b (40%)	2.4 ± 1.6 (67%)	23.7 ± 17.2 b (73%)
Dai et al., 2013 [36]	China	Chinese	14	22.9 ± 2.1	100	40	MD	14.45 ± 3.52 (24%)	90.38 ± 38.3 (42%)	5.22 ± 2.26 (43%)	53.3 ± 18.58 (35%)
Hasunuma et al., 2016 [31]	China	Chinese	40	31.5 ± 2.9	100	40	SD	3.33 ± 2.11 (63%)	14.3 ± 7.4 (52%)	1.16 ± 0.63 (54%)	13.6 ± 6.9 (51%)
Li et al., 2019 [37]	China	Chinese	60	28.5 ± 4.32	50	40	SD	6.93 ± 4.76 (69%)	30.18 ± 12.83 (43%)	3.22 ± 0.67 (21%)	21.02 ± 15.33 (6%)
							MD	11.24 ± 7.78 (69%)	42.95 ± 22.74 (53%)	4.66 ± 2.03 (44%)	34.72 ± 24.89 (72%)
Zhao et al., 2015 [38]	China	Chinese	15	25–59	N/A	20	SD	11.99 ± 6.68 (56%)	34.46 ± 16.38 c (48%)	2.41 ± 1.15 (48%)	18.24 ± 6.35 c (35%)
Hasunuma et al., 2016 [31]	Japan	Japanese	40	25.0 ± 4.0	100	20	SD	2.24 ± 1.46 (65%)	10.3 ± 6.3 (61%)	1.44 ± 0.78 (54%)	12.5 ± 6.0 (48%)
PK dataset excluded from meta‐analyses								Geometric mean ± geometric standard deviation (% geometric coefficient of variation)
Bergman et al., 2009 [39]	Belgium	Caucasians	12	21–40	75	20	SD	3.7 ± 4.6 (304%)	13.5 ± 9.9 (1381%)	0.8 ± 0.8 (23%)	8.2 ± 8.4 (958%)
Bernsdorf et al., 2006 [40]	Germany	Caucasians	18	21–30	56	40	MD	5.47 ± 0.81 (36%)	27.8 ± 0.68 b (70%)	0.35 ± 0.65 (81%)	32.4 ± 0.72 b (59%)
Birmingham et al., 2015 [17]	USA	Caucasians	30	32.9 ± 11.3	70	40	SD	8.63 ± 1.1 (10%)	32.2 ± 1.1 d (9%)	1.99 ± 1.12 (11%)	17.94 ± 1.12 d (11%)
Dingemanse et al., 2014 [41]	Belgium	Caucasians (97%) Asians (3%)	32	19–50	100	20	SD	2.915 ± 1.099 (9%)	10.330 ± 1.109 (10%)	1.165 ± 1.100 (10%)	11.300 ± 1.086 (8%)
Gehin et al., 2015 [42]	Germany	Caucasians (95%) Asians (5%)	22	34.7 ± 10.7	100	40	SD	7.84 ± 1.17 (16%)	31.22 ± 1.17 (16%)	1.57 ± 1.16 (15%)	15.83 ± 1.17 (16%)
Graefe‐Mody et al., 2010 [43]	Germany	Caucasians	20	26–58	100	40	MD	6.29 ± 1.98 (77%)	26.4 ± 1.92 b (73%)	1.64 ± 1.87 (69%)	17.5 ± 1.89 b (71%)
Hoch et al., 2013 (dataset 1) [44]	Germany	Caucasians (93%) Blacks (7%)	14	36.7 ± 8.1	100	40	SD	7.6 ± 1.3 (24%)	30.0 ± 1.3 (29%)	1.7 ± 1.1 (13%)	20.2 ± 1.2 (15%)
Hoch et al., 2013 (dataset 2) [45]	Germany	Caucasians	14	35.9 ± 7.4	100	40	SD	7.3 ± 1.2 (20%)	31.7 ± 1.2 (22%)	1.5 ± 1.2 (15%)	18.7 ± 1.2 (16%)
Birmingham et al., 2015 [17]	USA	Chinese	32	40.7 ± 16.9	53	40	SD	11.1 ± 1.1 (9%)	39.7 ± 1.1 d (9%)	2.95 ± 1.10 (9%)	22.97 ± 1.07 d (7%)
Zeng et al., 2022 [46]	China	Chinese	18	26.6 ± 6.0	100	20	SD	3.35 ± 1.67 (55%)	14.7 ± 1.59 (49%)	2.21 ± 1.60 (50%)	17.76 ± 1.54 (45%)
Birmingham et al., 2015 [17]	USA	Japanese	31	37.9 ± 14.0	74	40	SD	10.7 ± 1.1 (9%)	36.2 ± 1.1 d (9%)	3.23 ± 1.10 (9%)	24.22 ± 1.06 d (5%)

Abbreviations: MD, multiple dose; N/A, not applicable; SD, single dose.

^a Age are presented as arithmetic mean ± standard deviation or range.

^b AUC_0–24h.

^c AUC_0–48h.

^d AUC_0–t.

Our meta‐analysis of the SVA/SV C_max and AUC ratios showed significant differences across populations relative to unity (Figure 3A,B). The forest plots revealed that Thais exhibited markedly higher SVA relative to SV for both C_max and AUC, with mean SVA/SV ratios of 2.1 and 2.7, respectively. In contrast, Caucasians and Chinese showed substantially lower SVA relative to SV, with corresponding ratios of 0.2–0.4 and 0.7, respectively (Figure 3A,B). In Japanese, SVA C_max but not AUC values were lower than those of SV, with corresponding SVA/SV ratios of 0.6 and 1.2. Tests for subgroup differences indicated a statistically significant subgroup effect (p < 0.00001) for both C_max and AUC ratios. Further pairwise analyses confirmed statistically significant differences (p < 0.001) in both parameters for all Thai–non‐Thai comparisons (Table S4). Moderate to high heterogeneity (I ² > 50%) was evident, particularly for the C_max ratio in Chinese studies and for the AUC ratio across populations (Figure 3A,B).

FIGURE 3.

Forest plots for SVA‐to‐SV ratio for C_max (panel A) and AUC (panel B) in Thai, Caucasian, Chinese, and Japanese populations, following the oral administration of 20–60 mg SV. Each study's point estimate and 95% confidence interval are shown; pooled subgroup estimates (random‐effects) are displayed as diamonds. The vertical line marks a ratio of 1 (no difference). IV, inverse variance; SE, standard error.

3.3. Population‐Level Variants in SV PK‐Related Genes

The results revealed 14,841 variants in the SLCO1B1 and PON1/2/3 genes, of which 3,144 exhibited significant differences in AF between the two populations. To explain the observed increased SVA/SV AUC and C_max ratios in Thais vs. Caucasians, we looked for SLCO1B1 SNPs defining decreased or no function haplotypes (potentially increasing the plasma SVA levels) with a higher AF in Thais, and conversely SNPs defining increased‐function haplotypes per the CPIC guideline for statins (potentially reducing plasma levels of SVA through increasing its hepatic uptake) with a lower AF in Thais. For PON, decreased function SNPs with a lower AF in Thais than in Caucasians and increased function variants (with respect to statin hydrolysis) with a higher AF in Thais would be candidates of interest.

Notably, the SLCO1B1 missense variants rs11045819 (C>A) and rs34671512 (A>C), which are defined as increased‐function haplotypes *14 and *20, respectively [20], were significantly more frequent in Caucasians, with 50‐fold and ~3‐fold higher prevalence compared to Thais (Table 4). In our study, all 50 participants carried the reference allele for rs11045819, consistent with the very low AF (0.003) found with the Thai Genome Reference Database (Table 4), thus making this variant a potential causal candidate contributing to the observed ethnic PK difference between Thais and Caucasians. Moreover, in line with the Thai genome database, only three heterozygous carriers of rs34671512 were identified in the present study (Table 1). All three individuals exhibited significantly decreased AUC_0–12h values for SVA (17.83 vs. 32.19 ng·h/mL) but not SV (12.63 vs. 15.12 ng·h/mL) compared with the AA genotype (Figure S3A,B). Although there was a decreasing trend of the SVA/SV AUC ratio between the two groups, the difference was not significant (Figure S3C). Phenotypically, these three subjects are considered normal function (*1/*20, *20/*37), per the CPIC guideline, but increased function per Mykkanen et al. [47]. Nevertheless, considering the relatively small magnitude of the impact, it is uncertain whether this variant contributes meaningfully to the observed PK difference between Thais and Caucasians.

TABLE 4.

Allele frequency (AF) of candidate variants in Thai and Caucasian populations with expected functional consequences on plasma SVA levels.

SNP			AF a		Population‐level AF ratio (Thai: Caucasian)	Expected relative plasma SVA levels
rsID	Reported functional impact	Potential effect on SVA plasma levels	Thais ThaiGeR (current study)	Caucasians gnomAD
SLCO1B1 rs11045819	Increased SLCO1B1 function [20]	Decreased through increased hepatic uptake	0.0032 (0)	0.1590	1:50	Thais > Caucasians
SLCO1B1 rs34671512			0.0200 (0.03)	0.0536	1:2.7
SLCO1B1 rs4149056	Decreased SLCO1B1 function [20]	Increased through decreased hepatic uptake	0.1168 (0.09)	0.1587	1:1.4	Thais ≤ Caucasians
PON1 rs854571	Decreased statin hydrolysis [48]	Decreased	0.2241 (N/A)	0.7100	1:3.2	Thais > Caucasians
PON2 rs17876183	Decreased gene expression in liver tissues [26]	Decreased assuming gene expressions positively correlated with statin hydrolysis	0.0003 (N/A)	0.0225	1:75	Thais > Caucasians
PON3 rs149867961			0.0002 (N/A)	0.0285	1:142

Abbreviations: gnomAD, Genome Aggregation Database; N/A, not available; ThaiGeR, Thai Genome Reference Database.

^a AF values of Thais (current study) were sourced from all participants (n = 50)—see Table 1, while AF values of Caucasians were obtained from the NFE subset in gnomAD v3.1.2.

Interestingly, the AF of rs4149056 (c.521T>C), the key SNP defining no function haplotypes shown in this study and reported elsewhere [8, 9] to be associated with increased SVA levels, was found to be slightly lower in Thais than in Caucasians (Table 4). Therefore, the rs4149056 variant is unlikely to contribute to the PK difference observed in this study. For all SLCO1B1 variants, the eQTL analysis did not reveal any significant effects on gene expression.

Among the PON gene family, the PON1 rs854571 (A>G) SNP is associated with decreased atorvastatin hydrolysis [48]. It is considered a minor allele in Thais but a major allele in Caucasians (AF = 0.2241 vs. 0.7100) (Table 4). Additionally, the eQTL data showed that rs854571 was linked to decreased PON1 expression (NES = −0.314, Figure S4). Taken together, this variant may contribute partly to the elevated SVA/SV C_max and AUC ratios observed in Thais. Moreover, we found two promoter variants of PON2 (rs17876183) and PON3 (rs149867961), associated with decreased gene expression (NES = −0.880 and −0.684 for PON2 and PON3, respectively, Figure S4), and showing a 74–139‐fold higher AF in Caucasians than in Thais. However, there have been no reports regarding their functional impact on statin hydrolysis.

4. Discussion

In this study, we found that elderly subjects exhibited ~1.5‐fold higher SV C_max, whereas SV AUC remained unchanged. The elevated SV C_max suggests that the primary impact of age is at the pre‐systemic level and is consistent with expectations for oral administration of a high extraction ratio compound, such as SV, with decreased hepatic blood flow [49]. A decline in intestinal and hepatic blood flow with aging is well documented [50]. The relatively modest effect aligns with earlier estimates of an approximate 0.8% decline per year in hepatic clearance after the age of 40 years [51], as our elderly population (60–77 years old) represents the early elderly. The higher SV C_max without a corresponding change in AUC appeared inconsistent with the 2‐fold increases in both C_max and AUC reported for midazolam and atorvastatin in a previous study using a microdose [13]. Although the underlying reasons for the observed differences are unknown, it is worth noting that, unlike SV, midazolam and atorvastatin are classified as moderate clearance drugs (hepatic extraction ratio ~0.4 [52, 53]), whose clearances, and thus their oral PK profiles, would be influenced by changes in hepatic blood flow and intrinsic clearances. Interestingly, our finding that age affected only SV C_max but not AUC contrasts with an earlier study showing increases in C_max and AUC of total HMG‐CoA reductase inhibitors in elderly than in young patients [16]. This disparity may be partially related to the fact that the total HMG‐CoA reductase inhibitors comprise many components with potentially different PK characteristics than SV.

In the case of SVA, our results showed no age impact on its C_max and AUC, inconsistent with a previous study showing increased active HMG‐CoA reductase inhibitors in elderly compared to young patients [16]. The exact reason for this remains unknown but may partly reflect the presence of other active components (also detected as active HMG‐CoA reductase inhibitors) with PK characteristics distinct from SVA. The unchanged plasma exposure of SVA observed in this study may reflect a net neutral effect of age on its formation vs. elimination kinetics. Conceivably, age‐related reduction in hepatic blood flow could reduce SV clearance (and thus SVA formation and exposure) as well as SVA elimination (and therefore increase SVA exposure).

To our knowledge, this is the first study to report markedly elevated (~2–9‐fold) SVA exposure normalized to SV in Thais compared with Caucasians, Chinese, and Japanese, with SVA/SV C_max and AUC ratios of > 2.0 in Thais and ≤ 1.0 in these other populations. Notably, only one PK study in Japanese subjects was available, and the observed difference for this group should therefore be interpreted with caution. However, a separate analysis of available studies in Caucasians, Chinese, and Japanese that reported geometric means (Table 3) showed that their SVA/SV ratios were < 1.0 (Figure S5, Table S5), directionally consistent with those included in our analysis. Additionally, our finding of significantly higher (~2‐fold) SVA than SV levels corroborates an earlier PK report in Thai subjects [12]. A similar finding of nearly 10‐fold higher SVA than SV steady‐state level has also been observed in Thai patients (n = 89) who had been treated with SV for at least 2 weeks [9]. The higher SVA levels observed in our study were not likely due to ex vivo hydrolysis of SV, as care was taken to minimize this during sample handling, preparation, and analysis. Moreover, the higher normalized SVA levels in Thais than in Caucasians are in line with the earlier report showing a relatively high number of rhabdomyolysis cases in Thais receiving a lower SV dose compared with Caucasians [14]. Nevertheless, given the limited Thai sample size across two datasets (n = 68) and potential differences in sample handling and study designs, a confirmatory clinical SV PK study may be warranted in Thais vs. other ethnicities, particularly Caucasians, using consistent sample processing and analytical methods, as well as consistent clinical study designs, including SV formulations and dosing instructions, across populations.

Our pharmacogenetic analysis of polymorphic genes relevant to SVA disposition with well established (SLCO1B1) or at least shown (PON) functional impact on statins demonstrated that the AF of the well‐known SLCO1B1 rs4149056 (c.521 T>C) polymorphism was similar between Thais and Caucasians [17], and thus unlikely to contribute to the observed ethnic differences in SVA PK. A similar non‐causal relationship conclusion was reached for this variant regarding the apparent ethnic difference in SVA PK observed between Caucasians and Japanese/Chinese [17]. However, the marked differences between Thais and Caucasians in population‐level AF of the increased‐function SLCO1B1 missense variant rs11045819 (C>A) and the decreased function (regarding statin hydrolysis) PON1 variant rs854571 (A>G) aligned well with, and thus potentially a contributing factor to the higher SVA than SV levels observed in Thais vs. Caucasians. Beyond these variant candidates, we also found two promoter variants of PON2 (rs17876183) and PON3 (rs149867961), which are associated with decreased gene expression and exhibited > 50‐fold higher AF in Caucasians than Thais. However, their functional impact on SV hydrolysis, and thus their potential involvement in the observed ethnic differences in SVA PK, remains to be investigated. Other polymorphic genes that may alter SVA PK in the opposite direction to SV include SLCO1B3, CES1/2, and UGT1A1/3. However, we did not analyze these genes due to limited information on the functional impact of their variants on SV metabolism and the complex splicing of UGT1A genes [54]. Similarly, CYP3A4/5 polymorphisms were not considered, as their effects on SV and SVA kinetics would occur in the same direction, inconsistent with the elevated SVA/SV C_max and AUC ratios observed in Thais.

It is worth noting that the elderly subjects included in our study were healthy and mostly early elderly (60–77 years old), but many elderly patients taking SV have comorbidity, frailty and/or are late elderly (> 75 years old). Recent research combining population PK and physiologically based PK modeling demonstrated that a significant decrease in the oral clearance of amlodipine, a CYP3A substrate, was associated with the frailty phenotype in Chinese patients [55]. Thus, frail elderly patients receiving SV could experience a higher increase in SV C_max than that observed in the present healthy elderly. A similar outcome may also occur in late elderly patients, as their hepatic blood flow, and thus hepatic clearance, is expected to decline further [51]. Although aging impacts only SV, and not SVA, cautious dosing may still be warranted in elderly patients, particularly those with additional risk factors for statin‐associated toxicity.

Notably, our finding of higher normalized SVA levels in Thais than in Caucasians may help partly explain the previously reported higher incidence of rhabdomyolysis reported earlier in Thais [14]. The finding also supports the ICH E5 (R1) framework [56], which recommends conducting PK studies in new regulatory regions for medicines likely to be sensitive to ethnic factors. Such medicines include prodrugs, low oral bioavailability drugs, and highly metabolized compounds whose disposition kinetics depend on polymorphic enzymes—features present in SV. However, the ICH E5 guideline does not address polymorphic transporters, such as SLCO1B1, which is one of the likely contributing factors to the ethnic differences observed in this study. As such, ethnic factors may also need to be considered for active metabolites, especially those constituting the majority of a medicine's pharmacological activity. Whether these ethnic differences extend to atorvastatin and rosuvastatin PK between Thai and Caucasian populations and/or are related to AF differences in the polymorphic SLCO1B1 or PON variants among Asian populations remains to be investigated. Atorvastatin and rosuvastatin are known substrates of OATP1B1, and thus a similar effect of the SLCO1B1 rs11045819 (C>A) missense variant on their PK, as observed for SVA, is a theoretical possibility. However, unlike SV, atorvastatin and rosuvastatin are administered directly in their pharmacologically active acid forms, making them less sensitive to PON polymorphisms.

5. Conclusions

Aging altered SV PK, primarily increasing C_max, but had no significant effect on SVA exposure, following SV administration to healthy Thai subjects. Meta‐analyses demonstrated that Thais exhibit significantly higher SVA exposure than Caucasians, Chinese, and Japanese, consistent with the higher incidence of SV‐associated rhabdomyolysis reported in the Thai population. The pharmacogenetic analyses suggested that AF differences in SLCO1B1 rs11045819 and PON1 rs854571 variants, and possibly in promoter variants of PON2 (rs17876183) and PON3 (rs149867961), may partially explain the ethnic differences in SVA PK. Overall, our study highlights age and ethnicity as important determinants of SV and SVA PK, with implications for dosing considerations in Thai patients. These findings also provide a foundation for future studies to elucidate the role of polymorphic variants in individual and interethnic variability in statin PK.

Author Contributions

Thanate Srimatimanon: Performed the research, analyzed the data, wrote the manuscript. Udomsak Udomnilobol: Designed the research, performed the research, analyzed the data. Pajaree Chariyavilaskul, Sarawut Siwamogsatham, Kearkiat Praditpornsilpa, Yongkasem Vorasettakarnkij, Aisawan Petchlorlian, Natchaya Vanwong, Jittima Piriyapongsa, and Pavita Tipsombatboon: Performed the research. Varalee Yodsurang: Performed the research, wrote the manuscript. Thomayant Prueksaritanont: Designed the research, wrote the manuscript.

Funding

This study was funded by the Health Systems Research Institute (HSRI), Thailand, Grant no. 64–154.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: cts70472‐sup‐0001‐DataS1.docx.

Srimatimanon T., Udomnilobol U., Chariyavilaskul P., et al., “Intrinsic Factors Influencing Simvastatin and Simvastatin Acid Pharmacokinetics: Age‐Related Studies in Thai Adults and Cross‐Population Comparisons,” Clinical and Translational Science 19, no. 2 (2026): e70472, 10.1111/cts.70472.