0% READ
FULL TEXT
INTRODUCTION
The remarkable progress achieved in pharmacotherapy over the past century has transformed the management of acute and chronic diseases, substantially improving patient survival, quality of life, and healthcare outcomes. Despite these advances, considerable interindividual variability in drug response remains an important challenge in clinical medicine. Patients receiving the same medication at similar doses may experience markedly different outcomes, ranging from optimal therapeutic benefit to treatment failure or serious Adverse Drug Reactions (ADRs). Such variability reflects the complex interaction of genetic, physiological, environmental, pathological, and lifestyle factors that influence drug disposition and pharmacological response (Pirmohamed, 2023; Motulsky, 1957; Kalow, 2006; Evans and McLeod, 2003; Lauschke and Ingelman-Sundberg, 2026). The recognition that inherited genetic variation contributes substantially to this variability has fundamentally changed the understanding of drug therapy and established an important scientific basis for pharmacogenomics and precision medicine (Pirmohamed, 2023; Evans and McLeod, 2003; Lauschke and Ingelman-Sundberg, 2026).
The limitations of the traditional "one-size-fits-all" approach have become increasingly apparent with the growing complexity of modern pharmacotherapy. Conventional prescribing strategies are largely based on population averages and assume that patients with similar clinical characteristics will respond comparably to the same therapeutic regimen. In reality, substantial differences occur in drug absorption, distribution, metabolism, excretion, and pharmacodynamic response. Clinical factors such as age, sex, body composition, hepatic and renal function, comorbidities, concomitant medications, nutritional status, and environmental exposures contribute to this variability. In addition, inherited polymorphisms affecting drug-metabolizing enzymes, membrane transporters, receptors, signalling proteins, and immune-response pathways can substantially modify drug efficacy and toxicity (Pirmohamed, 2023; Evans and McLeod, 2003; Relling and Evans, 2015).
The clinical consequences of variability in drug response are particularly evident in the burden of ADRs. ADRs contribute substantially to morbidity, mortality, healthcare utilization, and treatment costs. Importantly, some serious ADRs are predictable and potentially preventable because they arise from inherited differences in drug metabolism or immune-mediated drug responses. Identification of clinically actionable pharmacogenomic variants therefore provides an opportunity to predict susceptibility to selected toxicities before treatment initiation and to guide drug selection or dose optimization (Pirmohamed, 2023; Phillips et al., 2001).
The conceptual origins of pharmacogenomics can be traced to the pioneering work of Arno Motulsky, who proposed in 1957 that inherited biochemical differences could explain individual variation in drug response. Werner Kalow subsequently contributed substantially to the understanding of inherited variation in drug metabolism, while Friedrich Vogel introduced the term pharmacogenetics in 1959. These landmark observations established the scientific foundation for investigating inherited variability in pharmacotherapy (Motulsky, 1957; Kalow, 2006; Vogel, 1959; Lander et al., 2001; Venter et al., 2001; Kalow, 2004). The completion of the Human Genome Project in 2001, together with advances in high-throughput genotyping, Genome-Wide Association Studies (GWAS), and next-generation sequencing, subsequently transformed the field from the study of individual genetic effects into the broader genome-wide discipline of pharmacogenomics (Lander et al., 2001; Venter et al., 2001; Kalow, 2004).
Fundamentals of Pharmacogenomics
Pharmacogenomics is the study of how genetic variation influences individual responses to medications. It integrates pharmacology, molecular genetics, genomics, bioinformatics, and clinical medicine to identify genetic determinants of drug efficacy, toxicity, and dosage requirements. Pharmacogenetics traditionally focused on individual genes or variants, whereas pharmacogenomics adopts a broader genome-wide perspective and considers interactions among multiple genetic and clinical determinants of drug response (Pirmohamed, 2023; Evans and McLeod, 2003; Relling and Evans, 2015; Caudle et al., 2017).
The transition from pharmacogenetics to pharmacogenomics has been facilitated by major advances in genomic technologies, including high-throughput genotyping, GWAS, Next-Generation Sequencing (NGS), Whole-Exome Sequencing (WES), and Whole-Genome Sequencing (WGS). These technologies have enabled identification of genetic variants associated with drug metabolism, transport, therapeutic targets, immune-mediated drug reactions, and treatment response. Integration of genomic information with resources such as Pharm GKB has further facilitated translation of genetic discoveries into clinically actionable recommendations (Thorn et al., 2013; Roden et al., 2011).
Drug response is a complex phenotype resulting from interactions among genetic, environmental, physiological, and disease-related factors. Pharmacokinetics describes drug absorption, distribution, metabolism, and excretion, whereas pharmacodynamics describes the interaction of a drug with its biological targets and the resulting therapeutic or toxic effects. Genetic variation can influence either or both pathways, producing substantial interindividual differences in drug exposure and clinical response (Evans and McLeod, 2003; Caudle et al., 2017; Ingelman-Sundberg, 2004).
Pharmacokinetic Variability
is strongly influenced by polymorphisms in genes encoding drug-metabolizing enzymes and membrane transport proteins. Cytochrome P450 (CYP450) enzymes metabolize a large proportion of clinically used medications. Variants in genes such as CYP2D6, CYP2C19, CYP2C9, CYP3A5, and CYP2B6 can alter enzyme activity and produce distinct metabolizer phenotypes. Reduced enzyme activity may increase systemic drug exposure and toxicity, whereas increased activity may reduce exposure and therapeutic efficacy (Ingelman-Sundberg, 2004; Meyer, 2000; Wilkinson, 2005).
Phase II metabolic pathways also contribute substantially to pharmacokinetic variability. Genetic variation in TPMT and NUDT15 can markedly alter thiopurine metabolism and increase the risk of severe myelosuppression. Similarly, reduced activity of DPYD, which encodes dihydropyrimidine dehydrogenase, increases exposure to fluoropyrimidines and predisposes susceptible patients to severe and potentially life-threatening toxicity (Relling et al., 2019; Amstutz et al., 2018).
Drug transport proteins provide another important source of pharmacokinetic variability. ATP-binding cassette and solute carrier transporters regulate drug absorption, distribution, hepatic uptake, and elimination. A well-established example is SLCO1B1, which encodes the hepatic transporter OATP1B1. The SLCO1B1 c.521T>C variant reduces hepatic uptake of simvastatin and increases systemic exposure and the risk of statin-associated myopathy (Niemi et al., 2011; Link et al., 2008).
Pharmacodynamic Variability
Pharmacodynamic genes influence drug effects by altering receptors, enzymes, ion channels, signaling pathways, or immune recognition. Consequently, individuals with comparable plasma drug concentrations may experience different therapeutic or toxic responses because of genetic differences in drug targets (Evans and McLeod, 2003; Caudle et al., 2017).
Clinically relevant examples include VKORC1, which contributes to warfarin dose variability, and somatic biomarkers such as EGFR, KRAS, and BRAF, which influence selection of targeted anticancer therapies. HLA variation also has major clinical importance because particular HLA alleles predispose genetically susceptible individuals to severe immune-mediated drug reactions (Relling and Klein, 2011; Whirl-Carrillo et al., 2021; Roden et al., 2019).
Genetic Polymorphisms and Functional Consequences
Genetic variation occurs in multiple forms, including single nucleotide variants, insertions, deletions, Copy Number Variations (CNVs), tandem repeats, and structural variants. Depending on their location and functional effect, variants may alter amino acid sequence, gene expression, messenger RNA splicing, protein stability, or enzyme activity (Thorn et al., 2013; Lauschke et al., 2024).
The clinical significance of a pharmacogenomic variant also depends on its population frequency. Important pharmacogenetic alleles are not distributed uniformly across populations. For example, CYP2C19 loss-of-function alleles are more frequent in several East Asian populations, while NUDT15 variants associated with thiopurine intolerance are particularly important in Asian populations. Population diversity therefore needs to be considered when interpreting pharmacogenomic findings and designing equitable precision-medicine strategies (Table 1) (Bush and Moore, 2012; Lauschke et al., 2022; Zhou et al., 2025).
Genetic Variability Influencing Drug Response
Interindividual variability in drug response is driven by inherited genetic differences that influence pharmacokinetic and pharmacodynamic processes. Environmental factors, disease states, age, sex, diet, and concomitant medications also contribute; however, genetic variation is an important determinant of differences in drug efficacy, toxicity, and dosage requirements (Pirmohamed, 2023; Evans and McLeod, 2003; Relling and Evans, 2015; Roden et al., 2011).
Single nucleotide variants represent one of the most common forms of genetic variation relevant to pharmacogenomics. They may influence protein structure, enzyme activity, transcriptional regulation, or messenger RNA stability. Many variants are clinically neutral, whereas others have sufficiently reproducible effects to be incorporated into pharmacogenomic guidelines (Caudle et al., 2017; Lauschke et al., 2024).
Structural variation is also important. CNVs are particularly relevant to CYP2D6, where gene deletion may result in absent enzyme activity and gene duplication or multiplication may contribute to ultrarapid metabolism. Accurate characterization of CYP2D6 can therefore require methods capable of detecting both sequence variation and copy-number changes.
Genetic variants influence drug response principally through pharmacokinetic and pharmacodynamic mechanisms. Pharmacokinetic variability changes systemic drug concentrations through effects on ADME processes, while pharmacodynamic variability alters target sensitivity, signalling, enzyme activity, or immune recognition (Evans and McLeod, 2003; Ingelman-Sundberg, 2004; Wilkinson, 2005).
The CYP450 family remains one of the most clinically important groups of pharmacogenes. Functional polymorphisms in CYP2D6, CYP2C19, CYP2C9, CYP3A5, and CYP2B6 produce distinct metabolizer phenotypes and have implications for several therapeutic classes, including antidepressants, opioids, antiplatelet agents, immunosuppressants, and anticoagulants (Ingelman-Sundberg, 2004; Meyer, 2000; Wilkinson, 2005; Relling and Klein, 2011).
Drug transporters also influence pharmacokinetic variability. Variants in SLCO1B1, ABCB1, and other transporter genes can alter drug disposition. The SLCO1B1 c.521T>C variant is a well-characterized example because reduced transporter activity increases simvastatin exposure and myopathy risk (Niemi et al., 2011; Link et al., 2008). Pharmacodynamic variability is particularly important for immune-mediated adverse reactions. HLA-B*57:01 is strongly associated with abacavir hypersensitivity, HLA-B*15:02 with carbamazepine-induced Stevens-Johnson syndrome/toxic epidermal necrolysis in susceptible populations, and HLA-B*58:01 with severe cutaneous adverse reactions to allopurinol (Mallal et al., 2002; Mallal et al., 2008; Chung et al., 2004; Saito et al., 2016).
Population-specific differences in pharmacogenomic variation are clinically important. Differences in allele frequencies and linkage patterns can affect the applicability of findings generated in one population to another. This emphasizes the need for genomic research that adequately represents diverse populations (Bush and Moore, 2012; Lauschke et al., 2022; Zhou et al., 2025). Overall, genetic variability provides the biological rationale for genotype-guided prescribing and supports the transition from empirical treatment toward individualized pharmacotherapy (Table 2).
Type of variation | Description | Example gene | Clinical significance |
|---|---|---|---|
Single nucleotide variant | Single-base substitution | CYP2C19 | Altered enzyme activity |
Copy number variation | Gene deletion or duplication | CYP2D6 | Poor or ultrarapid metabolism |
Insertion/deletion | Addition or loss of DNA bases | DPYD | Increased fluoropyrimidine toxicity |
Tandem repeat | Repeated DNA sequence | UGT1A1 | Altered glucuronidation |
HLA polymorphism | Immune-related genetic variation | HLA-B*57:01 | Drug hypersensitivity |
Major Pharmacogenes and their Clinical Significance
The successful implementation of pharmacogenomics depends on identification of pharmacogenes with reproducible and clinically actionable gene-drug associations. CPIC and DPWG have translated evidence for selected gene-drug pairs into prescribing recommendations, enabling clinicians to incorporate genotype information into therapeutic decision-making (Caudle et al., 2021; Swen et al., 2023).
Cytochrome P450 Enzyme Family
CYP450 enzymes constitute the most extensively studied group of pharmacogenes. Functional polymorphisms alter enzyme activity and produce predictable metabolizer phenotypes. CYP2D6 is involved in the metabolism of numerous antidepressants, antipsychotics, β-blockers, opioids, and other drugs. Structural variation and multiple star alleles contribute to substantial interindividual variability. CYP2D6 activity-score approaches help translate genotype information into clinically meaningful metabolizer phenotypes (Gaedigk et al., 2019; Gaedigk et al., 2008; Meyer, 2004).
CYP2C19 is particularly important in cardiovascular medicine because reduced-function alleles impair activation of clopidogrel and may increase the risk of adverse cardiovascular outcomes in susceptible patients (Pereira et al., 2020; Lee et al., 2022). CYP2C9 contributes to variability in warfarin and phenytoin metabolism, while CYP3A5 is particularly relevant to tacrolimus therapy in transplantation (Ingelman-Sundberg, 2004; Relling and Klein, 2011).
Phase II Drug-Metabolizing Enzymes
TPMT and NUDT15 are important determinants of thiopurine toxicity. Reduced activity increases the risk of severe myelosuppression, and genotype-guided dose reduction or alternative therapy can improve safety. DPYD encodes the rate-limiting enzyme involved in fluoropyrimidine catabolism. Reduced DPYD activity increases exposure to fluorouracil and capecitabine and can result in severe gastrointestinal, hematological, neurological, and mucocutaneous toxicity. Pre-treatment DPYD testing is therefore clinically important when fluoropyrimidine therapy is being considered (Amstutz et al., 2018; Henricks et al., 2018). UGT1A1 variation can influence irinotecan toxicity and may assist in identifying patients at increased risk of severe neutropenia (Innocenti et al., 2014).
Drug Transporter Genes
SLCO1B1 encodes OATP1B1 and is an established determinant of statin disposition. Reduced-function variants increase circulating statin concentrations and myopathy risk (Niemi et al., 2011; Link et al., 2008). Other transporters, including ABCB1, ABCG2, and SLC22A1, have been investigated in relation to numerous medications. However, evidence for routine clinical implementation is less consistent for many transporter gene-drug pairs than for established CYP, DPYD, TPMT, NUDT15, and HLA associations.
Human Leukocyte Antigen Genes
HLA variants influence immune-mediated drug reactions. HLA-B*57:01 screening before abacavir therapy is a well-established example of pharmacogenomic testing preventing severe hypersensitivity (Mallal et al., 2002; Mallal et al., 2008). HLA-B*15:02 testing before carbamazepine therapy is recommended for genetically susceptible populations because of the strong association with SJS/TEN (Chung et al., 2004). HLA-B*58:01 is associated with allopurinol-induced severe cutaneous adverse reactions and represents another important example of genotype-guided prevention of serious ADRs (Saito et al., 2016).
Emerging Pharmacogenes
Additional clinically relevant pharmacogenes include RYR1 and CACNA1S for malignant hyperthermia susceptibility and G6PD for susceptibility to hemolysis with selected oxidant drugs. Continued functional characterization of rare and structural variants is expected to expand the number of clinically actionable pharmacogenomic biomarkers (Lauschke et al., 2024; Ingelman-Sundberg et al., 2018).
Clinical Applications of Pharmacogenomics
Translation of pharmacogenomic discoveries into clinical practice enables individualized drug selection and dosing and may reduce preventable toxicity. The strongest clinical evidence currently exists for selected gene-drug pairs in cardiovascular medicine, oncology, infectious diseases, psychiatry, neurology, and transplantation (Caudle et al., 2021; Phillips et al., 2018).
Cardiovascular Medicine
Clopidogrel is one of the best-established cardiovascular applications. Because it is a prodrug requiring CYP2C19-mediated bioactivation, reduced-function CYP2C19 alleles can result in reduced active-metabolite formation and diminished platelet inhibition. Genotype-guided selection of alternative P2Y12 inhibitors has been investigated in prospective clinical trials, including TAILOR-PCI (Pereira et al., 2020; Lee et al., 2022).
Warfarin dosing is also influenced by CYP2C9 and VKORC1 variation. Incorporation of genetic information into dosing algorithms can improve prediction of dose requirements, although clinical implementation must consider patient-specific factors and available alternatives.
Oncology
DPYD testing can identify patients at increased risk of severe fluoropyrimidine toxicity. Prospective genotype-guided dose individualization has demonstrated that pre-treatment DPYD testing can reduce severe toxicity while maintaining treatment efficacy (Amstutz et al., 2018; Henricks et al., 2018). UGT1A1 genotype information may also contribute to individualized irinotecan therapy, particularly in patients at increased risk of neutropenia (Innocenti et al., 2014). In addition to germline pharmacogenomics, somatic biomarkers such as EGFR, ALK, KRAS, BRAF, and HER2 guide targeted cancer therapy and demonstrate the broader integration of genomic information into oncology.
Psychiatry and Neurology
CYP2D6 and CYP2C19 variants influence exposure to several antidepressants and other psychotropic medications.
Pharmacogenomic-guided prescribing has been associated with improved outcomes in selected patient populations, although the magnitude of clinical benefit varies among studies (Bousman et al., 2019; Brown et al., 2020). In neurology, HLA testing is particularly relevant to carbamazepine therapy in populations with a high prevalence of HLA-B*15:02 (Chung et al., 2004).
Infectious Diseases
HLA-B*57:01 testing before abacavir treatment is one of the most successful examples of pharmacogenomic implementation. Prospective screening substantially reduced clinically confirmed abacavir hypersensitivity (Mallal et al., 2002; Mallal et al., 2008).
Transplantation Medicine
CYP3A5 genotype influences tacrolimus metabolism. CYP3A5 expressers generally require higher doses than non-expressers to achieve therapeutic concentrations. Genotype-guided initial dosing may facilitate earlier achievement of target concentrations when combined with therapeutic drug monitoring.
Future Expansion
Pharmacogenomics is expanding into additional therapeutic areas as evidence accumulates and testing becomes more accessible. Pre-emptive testing, electronic health-record integration, and clinical decision support may facilitate broader implementation (Rim et al., 2025; van der Wouden et al., 2017) (Table 3).
Pharmacogenomic Guidelines, Databases, and Regulatory Landscape
Successful clinical implementation requires standardized evidence-based frameworks for translating genetic findings into prescribing recommendations. CPIC, DPWG, PharmGKB, PharmVar, regulatory agencies, and clinical decision-support systems have all contributed to this process (Caudle et al., 2021; Swen et al., 2023; Whirl-Carrillo et al., 2012; Gammal et al., 2024; Koopmann et al., 2022; Gaedigk et al., 2018; Sangkuhl et al., 2020).
Gene | Major drugs | Primary clinical consequence |
|---|---|---|
CYP2D6 | Codeine, tramadol, antidepressants | Altered metabolism and efficacy |
CYP2C19 | Clopidogrel, proton-pump inhibitors, selected antidepressants | Altered activation or clearance |
CYP2C9 | Warfarin, phenytoin, selected NSAIDs | Altered exposure and toxicity |
CYP3A5 | Tacrolimus | Individualized dose requirement |
TPMT | Azathioprine, mercaptopurine | Myelosuppression |
NUDT15 | Thiopurines | Severe leukopenia/myelosuppression |
DPYD | Fluorouracil, capecitabine | Severe fluoropyrimidine toxicity |
SLCO1B1 | Simvastatin | Statin-associated myopathy |
HLA-B*57:01 | Abacavir | Hypersensitivity |
HLA-B*15:02 | Carbamazepine | Stevens-Johnson syndrome/TEN |
HLA-B*58:01 | Allopurinol | Severe cutaneous adverse reactions |
Specialty | Pharmacogene | Drug(s) | Clinical implication |
|---|---|---|---|
Cardiology | CYP2C19 | Clopidogrel | Alternative antiplatelet therapy in reduced-function phenotypes |
Cardiology | CYP2C9, VKORC1 | Warfarin | Individualized dose selection |
Oncology | DPYD | Fluorouracil, capecitabine | Reduce severe toxicity |
Oncology | UGT1A1 | Irinotecan | Identify increased toxicity risk |
Oncology | TPMT, NUDT15 | Thiopurines | Prevent severe myelosuppression |
Psychiatry | CYP2D6, CYP2C19 | Antidepressants | Exposure-guided treatment |
Infectious diseases | HLA-B*57:01 | Abacavir | Prevent hypersensitivity |
Neurology | HLA-B*15:02 | Carbamazepine | Reduce SJS/TEN risk |
Transplantation | CYP3A5 | Tacrolimus | Individualized initial dosing |
Clinical Pharmacogenetics Implementation Consortium
CPIC develops evidence-based recommendations describing how available genotype information should be used to guide medication selection or dosing. Its guidelines cover numerous clinically relevant gene-drug pairs, including CYP2D6, CYP2C19, CYP2C9, CYP3A5, TPMT, NUDT15, DPYD, SLCO1B1, and selected HLA genes (Caudle et al., 2021; Gammal et al., 2024).
Dutch Pharmacogenetics Working Group
DPWG has developed pharmacogenomic guidelines that are widely used in Europe. Although CPIC and DPWG recommendations are generally concordant for many gene-drug pairs, differences can occur because of methodological and clinical considerations (Swen et al., 2023; Koopmann et al., 2022).
Pharmacogenomic Knowledgebases
PharmGKB provides curated information regarding gene-drug associations, clinical annotations, guidelines, and drug labels (Whirl-Carrillo et al., 2012). PharmVar provides standardized pharmacogene and star-allele nomenclature, while PharmCAT assists in converting genetic findings into CPIC-based clinical annotations (Gaedigk et al., 2018; Sangkuhl et al., 2020).
Regulatory Agencies and Drug Labelling
Regulatory agencies increasingly incorporate pharmacogenomic information into drug labelling. Such information may identify recommendations for genetic testing, dose modification, contraindications, or altered pharmacokinetics and pharmacodynamics.
Clinical Decision Support Systems
into electronic health records can automatically identify actionable gene-drug interactions and provide prescribing recommendations at the point of care. Such systems can facilitate consistent use of pharmacogenomic information and reduce interpretation errors (Hicks et al., 2016; Weitzel et al., 2024).
Global Harmonization
International harmonization of laboratory testing, nomenclature, phenotype classification, reporting, and clinical recommendations remains important for expanding pharmacogenomic implementation (van der Wouden et al., 2021) (Table 4).
Pharmacogenomic Testing Technologies
Clinical pharmacogenomics depends on accurate and reliable technologies capable of detecting relevant genetic variants. The appropriate testing method depends on the clinical question, required turnaround time, variant complexity, cost, and available laboratory infrastructure (van der Lee et al., 2020; Russell et al., 2021; Tafazoli et al., 2021).
PCR-Based Genotyping
PCR-based methods are widely used for predefined clinically actionable variants because of their speed, analytical accuracy, and relatively low cost. Their principal limitation is that they generally detect only variants included in the assay design and may not adequately identify novel variants or complex structural variation.
Microarray Platforms
Microarrays allow simultaneous interrogation of large numbers of predefined variants and are useful for pre-emptive pharmacogenomic screening. Their major limitation is restricted detection of variants not represented on the array.
Next-Generation Sequencing
NGS enables analysis of multiple pharmacogenes and can identify both common and rare variants. It is particularly useful for pharmacogenes with extensive allelic diversity. However, interpretation of rare variants, structural variation, and complex haplotypes requires sophisticated bioinformatics (Russell et al., 2021; Tafazoli et al., 2021; Enko et al., 2023).
Long-Read Sequencing
can improve characterization of structural variants, copy-number changes, and complex haplotypes that may be difficult to resolve with conventional short-read methods. Its current limitations include cost, technical complexity, and limited availability.
Multi-Omics Technologies
Future pharmacogenomic testing is likely to integrate genomic information with transcriptomics, epigenomics, proteomics, metabolomics, and other molecular datasets. These approaches may improve understanding of biological mechanisms underlying variable drug response, although many remain primarily research tools (Hasin et al., 2017; Karczewski and Snyder, 2018) (Table 5).
Artificial Intelligence and Digital Health in Pharmacogenomics
The expansion of genomic sequencing has generated large and complex datasets that require advanced analytical approaches.
Artificial Intelligence (AI), Machine Learning (ML), and digital-health technologies can assist in integrating genomic information with clinical variables and translating complex data into clinically useful information (Topol, 2019; Johnson et al., 2021; Wang et al., 2023).
Artificial Intelligence in Pharmacogenomic Data Analysis
AI can analyze large genomic and clinical datasets to identify patterns and associations relevant to drug response. Machine-learning models can incorporate genetic variants together with demographic, laboratory, disease, medication, and environmental information (Topol, 2019; Johnson et al., 2021; Wang et al., 2023).
Machine Learning for Predicting Drug Response
Supervised and unsupervised learning methods may assist in predicting drug response, toxicity, and treatment selection. Deep-learning approaches are particularly promising for integrating multimodal data, although clinical validation and interpretability remain essential before widespread adoption (Johnson et al., 2021; Choi et al., 2020).
Resource | Primary role | Major contribution |
|---|---|---|
CPIC | Clinical prescribing guidelines | Gene-specific therapeutic recommendations |
DPWG | Pharmacogenomic guidelines | Genotype-based prescribing |
PharmGKB | Knowledgebase | Clinical annotations and evidence |
PharmVar | Pharmacogene nomenclature | Standardized star-allele definitions |
PharmCAT | Clinical annotation software | Automated pharmacogenomic interpretation |
FDA | Regulatory authority | Pharmacogenomic drug labeling |
EMA | Regulatory authority | Biomarker-guided regulatory information |
Technology | Principle | Advantages | Limitations | Current use |
|---|---|---|---|---|
PCR | Detection of predefined variants | Rapid, inexpensive | Limited variant coverage | Routine testing |
Microarray | Multiplex predefined genotyping | High throughput | Cannot detect unrepresented variants | Pre-emptive testing |
Targeted NGS | Sequencing selected pharmacogenes | Common and rare variant detection | Bioinformatics requirements | Increasing clinical adoption |
WES | Coding-region sequencing | Broad variant detection | Limited regulatory coverage | Specialized use |
WGS | Whole-genome sequencing | Comprehensive coverage | Cost and interpretation | Advanced precision medicine |
Long-read sequencing | Long contiguous DNA reads | Structural variant/haplotype resolution | Higher cost and limited availability | Emerging |
Multi-omics | Integrated molecular datasets | Broader biological characterization | Limited clinical validation | Primarily research |
Clinical Decision Support Systems
Pharmacogenomic CDSS can retrieve stored genetic information when medications are prescribed and compare genotype information with evidence-based guidelines. Real-time alerts can recommend dose modification, alternative drugs, or additional monitoring (Hicks et al., 2016; Bell et al., 2014).
Electronic Health Records and Pre-emptive Pharmacogenomics
Because germline pharmacogenomic information generally remains stable throughout life, pre-emptive testing allows genetic information to be stored in electronic health records and reused whenever clinically relevant medications are prescribed (van der Wouden et al., 2017; Rasmussen-Torvik et al., 2014).
Emerging Digital Health Technologies
Cloud-based systems, secure genomic databases, federated learning, and other digital technologies may facilitate storage, sharing, and analysis of genomic data while addressing privacy and scalability challenges (Wang et al., 2023; Kairouz et al., 2021).
Future Perspectives
Important challenges for AI implementation include transparency, interpretability, external validation, data quality, cybersecurity, and regulatory oversight. AI should function as a clinical decision-support tool rather than a substitute for professional clinical judgment (Table 6).
Challenges and Limitations in Clinical Implementation
Despite substantial advances, clinical implementation remains inconsistent across healthcare systems. Scientific, technical, economic, educational, ethical, and regulatory barriers continue to limit routine use (van der Wouden et al., 2017; Verbelen et al., 2017; Dunnenberger and Cavallari, 2022).
Limited Clinical Evidence
Although many pharmacogenetic variants have been identified, only a subset has sufficient evidence for routine clinical implementation. Robust replication, multicenter studies, and clinical-outcome research remain necessary for many emerging gene-drug associations (Verbelen et al., 2017; Tuteja et al., 2022).
Ethnic Diversity and Population Representation
Underrepresentation of non-European populations in genomic research can limit the applicability of pharmacogenomic findings across diverse populations. Differences in allele frequencies and linkage patterns can influence the predictive performance of pharmacogenomic tests ((Petrovski and Goldstein, 2021; Popejoy and Fullerton, 2016).
Economic and Infrastructure Constraints
Testing costs, molecular diagnostic infrastructure, sequencing capacity, bioinformatics requirements, reimbursement policies, and availability of trained personnel can limit implementation, particularly in low-and middle-income settings (Veenstra et al., 2000; Plumpton et al., 2016).
Knowledge Gaps Among Healthcare Professionals
Limited pharmacogenomics education and confidence among healthcare professionals can impede appropriate interpretation and use of test results. Integration of pharmacogenomics into undergraduate and postgraduate education and continuing professional development is therefore important (Haga et al., 2012; Murphy et al., 2010).
Ethical, Legal, and Social Considerations
Pharmacogenomic testing raises concerns regarding privacy, informed consent, confidentiality, secondary use of genetic information, and potential discrimination. Appropriate governance and data-protection frameworks are necessary (Clayton et al., 2019; McGuire and Burke, 2011).
Interpretation of Rare and Novel Variants
NGS increasingly identifies rare and previously uncharacterized variants. Many lack sufficient evidence to determine their functional or clinical significance and may therefore remain variants of uncertain significance (Ingelman-Sundberg et al., 2018).
AI application | Function | Clinical significance |
|---|---|---|
Machine learning | Predicts drug response from genomic and clinical data | Personalized drug selection |
Deep learning | Identifies complex patterns | Precision oncology and prediction |
Clinical decision support | Provides prescribing recommendations | Reduced medication errors |
EHR integration | Stores and retrieves pharmacogenomic information | Pre-emptive prescribing |
Cloud computing | Genomic data storage and sharing | Multi-center integration |
Federated learning | Model development without centralized patient data | Improved privacy |
Blockchain | Secure data management | Genomic-data governance |
Challenge | Clinical impact | Potential solution |
|---|---|---|
Limited evidence | Restricts implementation | Multicenter prospective studies |
Population underrepresentation | Reduced applicability | Diverse genomic research |
High testing cost | Limited accessibility | Cost reduction and reimbursement |
Inadequate infrastructure | Delayed implementation | Molecular diagnostic investment |
Limited clinician knowledge | Poor utilization | Curriculum and continuing education |
Ethical/privacy concerns | Reduced public confidence | Governance and data protection |
Rare variants | Uncertain significance | Functional studies and databases |
Lack of standardization | Inconsistent recommendations | Harmonization and quality assurance |
Standardization and Quality Assurance
Variation among laboratories in assay design, variant coverage, nomenclature, phenotype assignment, and reporting can result in inconsistent clinical interpretation. Standardized nomenclature, laboratory quality assurance, and harmonized reporting are essential (Pratt et al., 2021).
Future Directions for Overcoming Barriers
Addressing these challenges requires collaboration among researchers, clinicians, pharmacists, geneticists, regulatory authorities, healthcare organizations, policymakers, and patients. Priorities include improving population representation, professional education, testing accessibility, reimbursement, laboratory infrastructure, and clinical decision-support integration (Table 7).
Pharmacogenomics in India: Current Status, Challenges, and Future Opportunities
Challenges, and Future Opportunities
India presents a particularly important setting for pharmacogenomics because of its extensive population diversity and the presence of numerous genetically distinct communities. Pharmacogenomic findings generated predominantly in European or other populations may not always be directly applicable to Indian populations (Ghosh and Sengupta, 2016; Indian Genome Variation Consortium, 2008; Thangaraj and Chaubey, 2009).
Genetic Diversity and Clinical Relevance
Variation in pharmacogenes including CYP2C19, CYP2C9, CYP2D6, CYP3A5, TPMT, NUDT15, SLCO1B1, and HLA genes has clinical implications for drug response in Indian populations. The distribution of clinically relevant variants may differ among population groups, emphasizing the need for representative Indian genomic datasets.
National Genomic Initiatives
The IndiGen initiative has generated a valuable resource of genetic variants from Indian genomes, while the GenomeIndia Project is developing a comprehensive catalogue of genetic diversity across Indian populations. These resources can support discovery and interpretation of population-specific pharmacogenetic variation (Jain et al., 2021; Bhattacharyya et al., 2025).
Clinical Implementation in India
Clinical pharmacogenomic implementation in India remains limited and is concentrated mainly in selected academic centres, tertiary hospitals, and specialized laboratories. International recommendations such as CPIC and DPWG therefore remain important resources while Indian-specific evidence and implementation frameworks continue to develop.
Challenges Specific to India
Important barriers include unequal access to molecular diagnostic facilities, testing costs, limited availability of trained personnel, variable healthcare infrastructure, and limited integration of pharmacogenomics into healthcare education and prescribing systems. Locally generated clinical and economic evidence will be important for informing implementation and policy.
Future Opportunities
Future priorities include development of Indian-specific pharmacogenomic guidelines, integration of pharmacogenomics into pharmacy and medical education, establishment of accredited laboratories, expansion of electronic health-record infrastructure, and incorporation of pharmacogenomic decision support into clinical workflows (Table 8).
Future Perspectives
Pharmacogenomics has evolved from an experimental discipline into an important component of precision medicine. Continued advances in genomic technologies, computational biology, AI, and systems medicine may transform pharmacogenomics from a reactive clinical tool into a proactive strategy for individualized treatment (Mathur and Sutton, 2017; van der Wouden et al., 2023; Wagner et al., 2021).
Expansion of Pre-emptive Testing
Pre-emptive pharmacogenomic testing allows relevant genetic information to be generated before medication exposure and stored for future clinical use. The PREPARE/U-PGx experience provides important evidence that pre-emptive pharmacogenomic testing can be implemented across healthcare systems (van der Wouden et al., 2023; Van Driest et al., 2014).
Domain | Current status | Future priority |
|---|---|---|
Population genomics | Expanding through national initiatives | Comprehensive population representation |
Clinical implementation | Limited to selected centers | Broader healthcare integration |
Guidelines | Predominantly international | Indian-specific recommendations |
Laboratory infrastructure | Uneven | Accredited molecular laboratories |
Professional education | Limited formal training | Curriculum integration |
Health informatics | Variable EHR adoption | Pharmacogenomic CDSS |
Research | Increasing genomic studies | Multicenter implementation trials |
Emerging area | Expected impact |
|---|---|
Pre-emptive pharmacogenomics | Lifelong availability of actionable genetic information |
Whole-genome sequencing | Comprehensive identification of genomic variation |
Long-read sequencing | Improved structural-variant and haplotype characterization |
Multi-omics | Enhanced prediction beyond genomics alone |
Artificial intelligence | Advanced predictive treatment models |
EHR integration | Automated real-time decision support |
Global genomic databases | Improved interpretation across populations |
Precision therapeutics | Integrated individualized treatment |
Whole-Genome Pharmacogenomics
Future testing may move beyond selected pharmacogenes toward comprehensive genomic profiling. Whole-genome sequencing can identify coding and non-coding variants, structural variation, and other genomic features that may contribute to drug response, although interpretation remains challenging (Wagner et al., 2021; Ashley, 2016).
Integration of Multi-Omics
Drug response is influenced by dynamic biological processes in addition to inherited genetic variation. Integration of genomics with transcriptomics, epigenomics, proteomics, metabolomics, microbiome data, and other molecular information may improve characterization of treatment response (Hasin et al., 2017; Karczewski and Snyder, 2018).
Artificial Intelligence and Predictive Precision Medicine
AI and ML may assist in integrating genomic information with EHR data, laboratory findings, imaging, and other clinical variables to predict treatment response and toxicity. Explainability, external validation, transparency, and regulatory oversight will remain essential (Rajkomar et al., 2019; Topol, 2019).
Global Collaboration and Standardization
International collaboration will be necessary to harmonize laboratory methodologies, nomenclature, phenotype classification, reporting, and clinical guidelines. Genomic databases representing diverse populations will improve variant interpretation and reduce disparities associated with population underrepresentation.
Toward Precision Therapeutics
The future of pharmacogenomics extends beyond prediction based solely on inherited variants toward integrated precision therapeutics. Genetic, molecular, clinical, environmental, and behavioral information may increasingly be combined to support individualized treatment decisions. Continued innovation, evidence generation, education, ethical governance, and equitable implementation will determine how successfully these technologies translate into routine patient care (Table 9).
CONCLUSION
Pharmacogenomics has emerged as an important component of precision medicine and has challenged the traditional one-size-fits-all approach to pharmacotherapy. By identifying genetic variation that influences drug metabolism, transport, target sensitivity, and immune-mediated toxicity, pharmacogenomics can support individualized drug selection and dosing and can prevent selected serious adverse drug reactions.
The strongest clinical applications currently involve well-established gene-drug pairs such as CYP2C19-clopidogrel, DPYD-fluoropyrimidines, TPMT/NUDT15-thiopurines, SLCO1B1-statins, CYP3A5-tacrolimus, and selected HLA-associated drug hypersensitivity reactions. Evidence-based recommendations from CPIC and DPWG, together with resources such as PharmGKB and PharmVar, have facilitated translation of genomic findings into clinical practice.
Technological advances in PCR-based testing, microarrays, NGS, long-read sequencing, electronic health records, clinical decision support, and artificial intelligence are expected to further expand the clinical utility of pharmacogenomics. At the same time, challenges involving evidence quality, population diversity, cost, infrastructure, professional education, privacy, rare-variant interpretation, and standardization must be addressed.
India represents an important opportunity because of its extensive genetic diversity and expanding national genomic initiatives. Development of representative Indian genomic resources and population-specific clinical evidence will be essential for equitable implementation.
Ultimately, pharmacogenomics should be viewed not as a replacement for clinical judgment but as an additional evidence-based tool for individualized therapeutic decision-making. Continued scientific innovation, rigorous clinical validation, multidisciplinary collaboration, and equitable implementation will be essential for realizing the potential of pharmacogenomics to improve medication safety, therapeutic effectiveness, and patient-centered care.
