Frequency of Key Pharmacogenes CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 in the Yogyakarta Population: Implications for Precision Prescribing in Indonesia

Abstract

Genetic variations in CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 significantly influence drug metabolism and transport, impacting therapeutic response and adverse event risk. While pharmacogenomic guidelines advocate for genotype‐guided therapy, no population‐specific frequency data exist for these pharmacogenes in Yogyakarta, Indonesia—a region where medication use and polypharmacy are common. This study aims to determine the allele and genotype frequencies of CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 in the Yogyakarta, Indonesia population to inform regionally appropriate prescribing practices. Genotyping was performed using probe‐specific real‐time PCR (RT‐PCR) platforms. The panel targeted clinically relevant variants across CYP2C9, CYP2C19, CYP2D6, and SLCO1B1, integrating SNP and copy number variant analysis where applicable. High frequencies of CYP2C19 *2/*3 and CYP2D6*10 alleles were observed, consistent with Southeast Asian profiles. SLCO1B1 reduced‐function variants were prevalent, suggesting elevated risk for statin‐induced myopathy. CYP2C9*3 allele was also detected, warranting caution in warfarin and NSAID prescribing. The observed genotype distributions highlight the need for region‐specific pharmacogenomic strategies in Indonesia. This foundational dataset supports the development of precision prescribing protocols and reinforces the importance of integrating genotyping into clinical practice to improve therapeutic safety and effectiveness.

Study Highlights.

WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?

CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 are clinically actionable pharmacogenes with established CPIC guidelines translating genotype into prescribing recommendations. These genes affect the metabolism and transport of widely used drugs, including NSAIDs, antiplatelets, antidepressants, and statins. However, Southeast Asian populations—particularly Indonesians—remain underrepresented in global pharmacogenomic databases, limiting region‐specific implementation of precision medicine.

WHAT QUESTION DID THIS STUDY ADDRESS?

This study examined allele frequencies and predicted metabolizer phenotypes of CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 in an adult cohort from Yogyakarta, Indonesia, and assessed the prevalence of CPIC‐defined actionable phenotypes in this previously underrepresented regional population.

WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?

The study provides novel pharmacogenomic data from Yogyakarta, contributing evidence from a predominantly Javanese population that has been largely absent from international PGx datasets. It quantifies the proportion of individuals with actionable phenotypes, helping to reduce the data gap in Indonesian and Southeast Asian populations.

HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?

These findings strengthen the evidence base for genotype‐guided prescribing in regional Indonesian settings. By generating data from an underrepresented population, this study supports more equitable precision medicine implementation and informs future multi‐ethnic pharmacogenomic research across Indonesia.

Pharmacogenomics is an emerging field with the potential to transform prescribing practices by identifying genetic factors that influence drug response, efficacy, and safety. ¹ , ² Among the most clinically actionable pharmacogenes, CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 play pivotal roles in the metabolism, activation, and transport of numerous medications used across various therapeutic domains. International guidelines such as CPIC and regulatory bodies have endorsed genotype‐guided interventions for drugs metabolized by these genes. ³ , ⁴ However, effective translation of such recommendations requires population‐specific data to inform clinical decision‐making.

To date, no comprehensive pharmacogenomic study has investigated the allele frequencies of these key pharmacogenes in the Yogyakarta region of Indonesia. This represents a critical gap, especially given the widespread use of medications in Indonesia—including agents with narrow therapeutic windows such as warfarin, clopidogrel, and simvastatin—and the increasing prevalence of polypharmacy, particularly among older adults and patients with chronic conditions. ⁵ , ⁶ Without regionally stratified genomic data, prescribing decisions risk being made without full consideration of gene–drug and drug–drug–gene interactions that could compromise therapeutic outcomes.

This study aims to characterize the allele frequencies of CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 among individuals in Yogyakarta, Indonesia. By establishing a foundational genotype distribution, we seek to support clinical implementation of pharmacogenomic screening and promote safer, more effective prescribing practices tailored to the genetic landscape of the local population.

METHODS

Study design and population

This study aimed to examine the distribution of pharmacogenetic variants in CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 among individuals from Yogyakarta, Indonesia. Participants were Indonesian citizens aged 18 years or older, recruited from Dr. Sardjito General Hospital and Universitas Gadjah Mada Academic Hospital (RSA UGM). Accordingly, the study population represents a hospital‐based adult cohort. Sampling was performed purposively based on willingness to participate. Recruitment commenced after ethical clearance was granted by the UGM Medical and Health Research Ethics Committee (No. KE/FK/0992/EC/8 Juli 2024). Written informed consent was obtained from all participants prior to blood collection, including specific consent for genetic testing and pharmacogenomic analysis.

Sample collection and DNA extraction

Peripheral venous blood samples were collected by trained healthcare professionals. Sample were processed and stored at −20°C. Genomic DNA was extracted using the FavorPrep Genomic DNA Mini Kit (FavorGen), and quality was assessed using a Maestro spectrophotometer, targeting a 260/280 absorbance ratio between 1.8 and 2.0.

Genotyping and phenotype assignment

Genotyping was performed using the Nala PGx Core® panel, a real‐time PCR–based assay (Roche LightCycler 480 and Bio‐Rad CFX Opus 96 Dx Touch platforms) targeting clinically relevant variants in CYP2C9 (*2, *3), CYP2C19 (*2, *3, *17), CYP2D6 (*2, *3, *4, *5 [gene deletion], *6, *8, *9, *10, *14, *21, *29, *31, *35, *36, *41), and SLCO1B1 (rs4149056). For CYP2D6, copy number variation (CNV) assays targeting Intron 2 and Exon 9 were included to detect gene deletions, duplications, and hybrid alleles. ⁷ Laboratory procedures followed standardized protocols with internal quality control measures. Amplification performance and allelic discrimination plots were reviewed to ensure analytical validity. Genotyping was performed in a research laboratory setting under standardized quality‐controlled procedures.

The analytical performance of the Nala PGx Core® platform has been previously validated. The assay requires approximately 2 ng/μL genomic DNA. The assay demonstrated > 96% variant‐level concordance with benchmark methods across the evaluated targeted variants, with high diplotype‐level agreement. Robust precision was also reported for CYP2D6 copy number variation analysis. Overall, the platform shows high analytical sensitivity and reliability for all targeted variants included within the panel design. ⁷

Genotypes were translated into star (*) alleles according to predefined allele definitions of the Nala PGx Core™ panel, which align with CPIC star allele nomenclature. For CYP2D6, metabolizer status was determined using the activity score (AS) framework, where allele activity values (normal = 1.0; decreased = 0.5; no function = 0) were summed and adjusted for gene copy number. Activity score cut‐offs were applied as follows: AS = 0 (poor metabolizer), AS 0.25–1.0 (intermediate metabolizer), AS 1.25–2.25 (normal metabolizer), and AS > 2.25 (ultrarapid metabolizer), consistent with CPIC recommendations. ² Diplotypes were assigned by integrating SNP/indel results with CYP2D6 copy number variation (CNV) data. Phenotype classification for CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 followed CPIC and international pharmacogenetics consortium recommendations and was generated using the Nala Clinical Decision Support™ system. ⁷

Statistical analysis

Participant characteristics were summarized descriptively. Categorical variables are presented as frequencies and percentages. Hardy–Weinberg equilibrium (HWE) for CYP2C9, CYP2C19, and SLCO1B1 was assessed using the chi‐square (χ ²) goodness‐of‐fit test by comparing observed and expected genotype frequencies. CYP2D6 was not evaluated for HWE because its complex structural variation, including gene deletions and duplications, violates HWE assumptions. Exact 95% confidence intervals for allele frequencies were calculated using the Clopper–Pearson method with an online calculator (GraphPad QuickCalcs ((https://www.graphpad.com/quickcalcs/confinterval1/).)).

RESULT

The cohort included 190 participants, predominantly female (57.37%). The largest age groups were 20–29 years (20.53%) and 60–69 years (20.00%), with smaller proportions in other age categories and 10.00% aged ≥ 70 years. Most participants were Javanese (78.95%), followed by mixed Javanese ancestry (12.11%), while Chinese (2.11%), Sundanese (1.58%), and other ethnicities (5.26%) comprised minority groups.

Among the 190 participants, CYP2C9 was predominantly normal metabolizer (*1/*1, 95.26%), with smaller proportions of intermediate (*1/*3, 4.21%) and poor metabolizers (*3/*3, 0.53%). For CYP2C19, normal metabolizers comprised 63.16%, followed by intermediate metabolizers (*1/*2 and *1/*3, 29.47%), while rapid (*1/*17) and poor (*2/*2) metabolizers accounted for 3.16% and 4.21%, respectively. CYP2D6 demonstrated substantial genetic diversity, with intermediate metabolizers constituting the largest proportion of the cohort. For SLCO1B1, most participants had normal function (TT, 82.63%), with 16.32% showing decreased function (TC) and 1.05% poor function (CC) (Figure 1).

Figure 1.

Genotype and phenotype distribution of pharmacogenes in the study cohort (n = 190). (a) CYP2C9 genotype distribution showing predominance of normal metabolizers (*1/*1). (b) CYP2C19 genotype distribution with normal (*1/*1) and intermediate metabolizers (*1/*2, *1/*3) as the most frequent groups. (c) CYP2D6 genotype‐level distribution incorporating copy number variation (CNV), with phenotype classification based on CPIC activity score criteria. (d) SLCO1B1 genotype distribution demonstrating predominance of normal function (TT) followed by decreased function (TC) and poor function (CC). Percentages represent proportions within the total cohort.

Allele‐level analysis of CYP2C9 showed a predominance of the *1 allele (97.37%; 95% CI: 95.21–98.73), while the *3 allele was infrequent (2.63%; 95% CI: 1.27–4.79) (Table 1). Genotype distributions showed minor deviation from Hardy–Weinberg equilibrium (p < 0.05), likely reflecting hospital‐based sampling and the low frequency of the *3 allele rather than genotyping error; all samples passed internal quality control.

Table 1.

CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 genotype distribution

Gene/Allele	Frequency (%)	95% CI (Clopper–Pearson)
CYP2C9
*1	97.37	95.21–98.73
*3	2.63	1.27–4.79
CYP2C19
*1	79.47	75.06–83.42
*2	17.63	13.93–21.84
*17	1.58	0.58–3.40
*3	1.32	0.43–3.04
SLCO1B1
T	90.79	87.42–93.50
C	9.21	6.50–12.58
CYP2D6
*1	25.00	20.8–29.7
*2	7.11	4.8–10.2
*4	1.05	0.29–2.65
*5	4.74	2.8–7.4
*10	37.11	32.2–42.2
*36	19.47	15.6–23.8
*39	0.26	0.01–1.45
*41	5.26	3.3–8.0
Copy number ≥3	41.58	34.5–48.9

For CYP2C19, the *1 allele was most common (79.47%; 95% CI: 75.06–83.42), followed by the loss‐of‐function *2 allele (17.63%; 95% CI: 13.93–21.84). The *17 allele (1.58%; 95% CI: 0.58–3.40) and the *3 allele (1.32%; 95% CI: 0.43–3.04) were observed at low frequencies, suggesting that reduced CYP2C19 activity in this population is primarily driven by the *2 allele. Genotype distributions conformed to Hardy–Weinberg equilibrium (p > 0.05) (Table 1).

For CYP2D6, substantial allelic diversity was observed. The decreased function *10 allele predominated (37.11%; 95% CI: 32.2–42.2), followed by *1 (25.00%; 95% CI: 20.8–29.7) and *36 (19.47%; 95% CI: 15.6–23.8). Copy number variation was frequent, with 41.58% of participants carrying CYP2D6 copy number ≥ 3 (95% CI: 34.5–48.9) (Table 1).

For SLCO1B1, the T allele was predominant (90.79%; 95% CI: 87.42–93.50), while the decreased function C allele accounted for 9.21% (95% CI: 6.50–12.58). Genotype distributions conformed to Hardy–Weinberg equilibrium (Table 1).

DISCUSSION

This study focused on four key pharmacogenes identified by the Clinical Pharmacogenetics Implementation Consortium (CPIC) as clinically actionable. ² Robust CPIC guidelines translate genotypes of CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 into prescribing recommendations. These genes affect the metabolism and transport of commonly used drugs, including NSAIDs, antiplatelets, antidepressants, and statins. ⁵ , ⁶ Characterizing their allele frequencies and predicted phenotypes in the Yogyakarta, Indonesian population is crucial for optimizing drug safety and efficacy, particularly given the limited population‐specific data available.

The frequency of the reduced‐function allele CYP2C9*3 in our Indonesian cohort was consistent with most Southeast Asian populations, though higher frequencies occur in some Malaysian Indian groups. ⁸ This allele significantly reduces enzymatic activity and is associated with slower metabolism of drugs like warfarin and phenytoin. ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² In Indonesia, where warfarin is widely used for conditions such as atrial fibrillation, ⁹ the presence of CYP2C9*3 suggests a notable subset of patients may require lower maintenance doses and closer INR monitoring to prevent bleeding events. While overall similarity to regional data supports the applicability of existing genotype‐guided dosing recommendations, observed subgroup variability highlights ethnic heterogeneity within Southeast Asia and underscores the need for population‐specific algorithms where feasible.

Compared to regional Southeast Asian data, our cohort showed lower CYP2C19*2 frequencies than reported for Singapore Chinese, Malay, and Indian populations. The CYP2C19*3 allele (1.3%) was also lower than most regional groups, while CYP2C19*17 (1.6%) was markedly lower than in Singapore Indians (15.1–21.4%) ⁸ , ¹⁰ . These discrepancies reflect population stratification within Indonesia and underscore the need for population‐specific data.

The high CYP2C19*1 frequency suggests a lower population‐level likelihood of reduced CYP2C19 activity. However, 64 participants (33.7%) carried genotypes associated with intermediate or poor metabolizer phenotypes (e.g., *1/*2, *1/*3, *2/*2), for which CPIC guidelines recommend alternative antiplatelet strategies due to diminished clopidogrel activation and increased therapeutic failure risk.^13,18 In Indonesia, where clopidogrel is commonly prescribed post‐PCI, genotype‐guided selection of alternatives (e.g., ticagrelor, prasugrel) could reduce thrombotic events in this substantial at‐risk subgroup, consistent with evidence that genotype‐guided prescribing improves cardiovascular outcomes. ¹¹

Comparison with previously published Southeast Asian datasets indicates that the CYP2D6 allele distribution in the Yogyakarta cohort is consistent with regional patterns. The high frequency of CYP2D6*10 (37.11%) falls within the range reported in Chinese, Thai, and Singaporean cohorts, while the CYP2D6*36 allele (19.47%) is also comparable to frequencies observed in several Southeast Asian populations. In contrast, the functional CYP2D6*1 allele (25%) appears relatively lower, reflecting the predominance of reduced‐function alleles in this cohort.

Clinically, these findings are particularly relevant in Indonesia, where CYP2D6 substrates such as tramadol and antidepressants are widely prescribed. ¹² , ¹³ The high prevalence of reduced‐function alleles suggests an increased likelihood of reduced prodrug activation and altered drug exposure, underscoring the potential value of genotype‐guided prescribing in routine clinical practice.

In our study, 17.37% of individuals carried SLCO1B1 decreased or poor function genotypes (TC or CC). The rs4149056 C allele frequency in our cohort is broadly consistent with previously reported Southeast Asian populations, including Thai, Chinese, and Malay cohorts, and remains lower than frequencies typically observed in European populations ⁷ . These findings support regional consistency in SLCO1B1 variant distribution while highlighting interethnic variability across global populations.

Given that OATP1B1 mediates hepatic uptake of statins, reduced‐function variants increase systemic statin exposure and the risk of statin‐associated musculoskeletal symptoms. ¹⁴ , ¹⁵ Although the prevalence of risk alleles in our population is moderate relative to other global populations, the presence of decreased function variants in nearly one in six individuals remains clinically meaningful. In Indonesia, where statins are widely prescribed ¹⁶ , SLCO1B1 genotyping may support individualized statin selection or dose adjustment in accordance with CPIC recommendations.

This study has several limitations. Participants were recruited from hospital‐based institutions in Yogyakarta and consisted predominantly of individuals of Javanese ancestry; therefore, the findings may reflect regional allele distributions rather than the full ethnic diversity of Indonesia. Given the marked genetic heterogeneity across Indonesian subpopulations, extrapolation to other ethnic groups should be made cautiously. Hospital‐based recruitment may also represent healthcare‐seeking individuals rather than a strictly population‐random sample.

Although the sample size is comparable to prior Southeast Asian pharmacogenomic studies, larger multi‐center investigations including diverse Indonesian ethnic groups would enable more precise frequency estimates and improved stratification. Furthermore, while the genotyping platform captures clinically actionable variants and copy number variation, rare or population‐specific variants outside the assay design may not have been detected, potentially resulting in modest overestimation of reference (*1) alleles.

Despite these limitations, pharmacogenomic data from well‐characterized Indonesian subpopulations remain limited. This study provides foundational allele frequency data to support future nationwide, multi‐ethnic studies incorporating comprehensive sequencing and clinical outcome data to guide genotype‐informed prescribing in Indonesia.

Future prospective studies with larger, outcome‐linked cohorts are needed to assess genotype‐guided dose requirements, drug response, and adverse event risk across age, sex, and ethnic subgroups in Indonesian patients.

CONCLUSION

Our findings demonstrate that genotyping of CYP2C9, CYP2C19, CYP2D6, and SLCO1B1 characterizes clinically actionable pharmacogenomic variation within a Yogyakarta‐based cohort. By describing population‐specific allele and predicted phenotype distributions, this study provides regionally relevant data to inform genotype‐guided prescribing. These results support the integration of pharmacogenomic screening into clinical workflows in underrepresented populations, while acknowledging that clinical outcome studies are needed to further define its impact beyond single‐gene associations.

CONFLICTS OF INTEREST

The authors declared no competing interests for this work.

AUTHOR CONTRIBUTIONS

DAAN wrote the manuscript; DAAN, SC, WRP, M, and DSM designed the research; SC, WR, DSM, RAW, M, and DAAN performed the research; DAAN, RAW, and SC analyzed the data.

ETHICS STATMENT

Ethical clearance was granted by the UGM Medical and Health Research Ethics Committee (No. KE/FK/0992/EC/8 Juli 2024).

CONSENT

Not applicable.

DATA AVAILABILITY STATEMENT

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.