PHARMACOGENOMICS IN RARE DISEASE MANAGEMENT: A Comprehensive Review of Principles, Applications, and Emerging Frontiers

The global burden of rare diseases represents one of medicine's most persistent and underappreciated challenges. Although each individual rare disease affects only a small fraction of the population, the cumulative impact across the more than 7,000 recognized rare conditions is enormous: the National Organization for Rare Disorders estimates that approximately 30 million Americans and 300 million individuals worldwide live with a rare disease, and the European Commission places a similar figure across EU member states (NORD, 2023; Ferreira, 2019). In the United States, a rare disease is officially defined as one affecting fewer than 200,000 Americans; the European Union employs a prevalence threshold of 1 in 2,000 individuals, while other jurisdictions utilize different but broadly comparable thresholds (Orphanet, 2023). A defining characteristic of rare diseases is their predominantly monogenic etiology. Approximately 80% of recognized rare conditions have an identifiable genetic basis, most commonly involving pathogenic variants in single genes that disrupt critical biosynthetic, metabolic, structural, or regulatory pathways (Hamosh et al., 2005). This genetic architecture makes rare diseases uniquely amenable to pharmacogenomic investigation: when the molecular basis of a disease is known, it becomes possible, in principle, to predict how that molecular defect will interact with therapeutic agents, to identify patient subgroups most likely to benefit from specific interventions, and to anticipate adverse drug reactions that arise from genetically determined alterations in drug metabolism or target biology. Pharmacogenomics—the study of how an individual's complete genomic profile influences drug response—has its conceptual roots in mid-twentieth-century observations about heritable differences in drug metabolism. The field has since matured into a sophisticated discipline encompassing the analysis of pharmacokinetic variants in drug-metabolizing enzymes and transporters, pharmacodynamic variants in drug targets, and genome-wide association studies of complex drug response phenotypes (Relling & Evans, 2015). The Clinical Pharmacogenetics Implementation Consortium (CPIC) and the Dutch Pharmacogenetics Working Group (DPWG) have formalized many pharmacogenomic insights into actionable clinical guidelines that are increasingly integrated into electronic health record systems (Caudle et al., 2014; Swen et al., 2011). Despite these advances in mainstream pharmacogenomics, the application of pharmacogenomic principles specifically to rare disease management has received comparatively limited systematic attention. Several factors account for this gap. Small patient populations make conventional clinical trial designs statistically underpowered. The orphan drug development pathway, while providing important regulatory incentives, has historically prioritized proof-of-concept over pharmacogenomic characterization. The rarity of individual conditions has impeded the accumulation of the large genomic and clinical datasets necessary for robust pharmacogenomic discovery. Advances in variant interpretation, driven by growing databases such as ClinVar and databases of gnomAD population variants, are improving the clinical actionability of genomic information (Landrum et al., 2014). Gene therapy and RNA-based therapeutic modalities are creating entirely new classes of treatments whose application is inherently pharmacogenomic. And the increasing sophistication of rare disease registries and natural history studies is providing the longitudinal clinical data needed to link genotypic variation to drug response phenotypes. This review aims to provide a comprehensive, authoritative synthesis of the current state of pharmacogenomics in rare disease management. We examine foundational concepts in pharmacogenomics and rare disease biology, survey disease-specific applications across major rare disease categories, analyze the technological platforms enabling pharmacogenomic discovery and implementation, discuss frameworks for clinical translation and implementation, and identify key ethical, regulatory, and equity challenges. We conclude by outlining the most promising emerging directions that will shape the future of precision medicine for rare disease patients.

1. Foundational Principles of Pharmacogenomics

2.1 Historical Development and Core Concepts

The intellectual lineage of pharmacogenomics extends to the 1950s and 1960s, when clinicians first documented heritable differences in drug response. Arno Motulsky's 1957 observation that adverse reactions to certain drugs—including primaquine-induced hemolysis in glucose-6-phosphate dehydrogenase (G6PD)-deficient individuals—had a genetic basis laid the conceptual groundwork for the field (Mancinelli, Cronin, & Sadée, 2000). Subsequent decades brought the characterization of slow and rapid metabolizer phenotypes for isoniazid (driven by N-acetyltransferase 2 polymorphisms), the molecular cloning of the cytochrome P450 (CYP) enzyme superfamily, and the recognition that common variants in these enzymes explained much of the population-level variability in drug pharmacokinetics. The term 'pharmacogenomics' gained currency in the late 1990s as genome-wide methodologies began to supersede the candidate gene approaches of earlier pharmacogenetics. Whereas pharmacogenetics traditionally focused on single-gene variants affecting specific drug pathways, pharmacogenomics embraces the entire genome as the unit of analysis, encompassing variants in drug-metabolizing enzymes, drug transporters, drug targets, immune response genes, and genes modifying disease susceptibility and progression (Langreth & Waldholz, 1999; Collins & Varmus, 2015). This broader scope is particularly relevant in rare disease management, where the disease-causing variant itself often has direct pharmacological implications that intersect with secondary pharmacokinetic variation in metabolizing enzymes.

2.2 Pharmacokinetic Pharmacogenomics: Drug-Metabolizing Enzymes

The cytochrome P450 (CYP) enzyme superfamily is the dominant system for the oxidative metabolism of the majority of clinically used drugs. Among the more than 50 human CYP enzymes, CYP2D6, CYP2C19, CYP2C9, CYP3A4, and CYP3A5 collectively account for the metabolism of approximately 60 to 70 percent of all prescribed medications (Zanger & Schwab, 2013). Each of these enzymes is encoded by highly polymorphic genes, with numerous single-nucleotide polymorphisms (SNPs), insertions, deletions, gene duplications, and copy number variants generating a spectrum of metabolizer phenotypes ranging from poor metabolizers (who carry two non-functional alleles and therefore cannot metabolize substrates efficiently) to ultrarapid metabolizers (who carry gene duplications and therefore metabolize substrates at dramatically elevated rates). The clinical consequences of these metabolizer phenotypes are profound and clinically actionable. Poor metabolizers of a pro-drug that requires CYP activation may experience complete therapeutic failure; ultrarapid metabolizers of an active drug may experience sub-therapeutic exposure and lack of efficacy; and poor metabolizers of drugs with narrow therapeutic indices may accumulate toxic plasma concentrations even at standard doses (Ingelman-Sundberg, 2004). CPIC has published comprehensive prescribing guidelines for numerous drug-gene pairs involving CYP2D6, CYP2C19, CYP2C9, and other enzymes, providing clinicians with genotype-specific dosing recommendations that have been validated in prospective clinical studies (Caudle et al., 2014; Lee et al., 2020).

2.3 Pharmacodynamic Pharmacogenomics: Drug Targets and Modifiers

Pharmacodynamic pharmacogenomics examines how genetic variation in drug targets and downstream signaling pathways influences the magnitude and character of drug response at the receptor or molecular level, independent of pharmacokinetic considerations. In the context of rare diseases, pharmacodynamic pharmacogenomics is particularly illuminating because the disease-causing mutation often directly resides within the drug target or its functional pathway. This creates a situation in which the pharmacogenomic characterization of the patient is inseparable from the diagnosis of the disease itself. The canonical example of this phenomenon is cystic fibrosis (CF), where over 2,000 pathogenic variants in the CFTR gene have been catalogued, and where the clinical efficacy of CFTR modulator drugs is strictly dependent on the specific variant class present in an individual patient. Ivacaftor, the first approved CFTR potentiator, is effective only in patients carrying specific gating mutations (most notably G551D); it produces no meaningful clinical benefit in patients homozygous for the most common CFTR variant, F508del, because the F508del protein does not traffic to the cell surface where ivacaftor can act (Cystic Fibrosis Foundation, 2023). This variant-specific response pattern represents pharmacodynamic pharmacogenomics in its most direct and therapeutically consequential form.

2. Genetic Architecture of Rare Diseases and Pharmacogenomic Implications

3.1 Mutational Diversity and Allelic Heterogeneity

A fundamental genomic characteristic of rare diseases is extreme allelic heterogeneity—the existence within a single disease gene of a very large number of distinct pathogenic variants, each of which may produce subtly or dramatically different disease phenotypes and carry different implications for therapeutic response. The Online Mendelian Inheritance in Man (OMIM) database catalogs over 7,000 phenotypes with a known molecular basis, with the number of distinct causative variants in individual disease genes ranging from a handful to several thousand (Hamosh et al., 2005). This allelic diversity has profound implications for pharmacogenomics: therapeutic strategies that are effective for one variant may be completely ineffective—or even harmful—for another, necessitating precise molecular diagnosis before therapeutic decisions are made. The molecular consequences of pathogenic variants span a spectrum from complete loss of function (null alleles) through partial loss of function to gain-of-function or dominant-negative effects. This functional spectrum is directly relevant to pharmacogenomic drug selection. For example, enzyme replacement therapies (ERTs) in lysosomal storage disorders are generally most effective when residual enzyme activity is absent or very low; patients with partial enzyme deficiency due to missense variants may have different treatment trajectories than those with complete absence of enzyme due to nonsense mutations or large deletions (Desnick & Schuchman, 2012).

3.2 Genotype-Phenotype Correlations

Establishing robust genotype-phenotype correlations is a prerequisite for meaningful pharmacogenomic stratification in rare diseases. When specific genotypes reliably predict specific phenotypic presentations—including disease severity, rate of progression, organ involvement, and likelihood of response to particular therapies—clinicians can use molecular diagnosis not only to confirm a diagnosis but also to guide therapeutic decision-making prospectively. Such correlations have been established with varying degrees of robustness across different rare disease areas. In Gaucher disease, the most prevalent lysosomal storage disorder, the N370S/N370S genotype of the GBA gene is associated with a predominantly non-neuronopathic (type 1) phenotype and typically a good response to enzyme replacement therapy with imiglucerase or velaglucerase alfa, whereas the L444P allele, particularly in compound heterozygosity or homozygosity, correlates with more severe disease including neurological involvement and more variable treatment response (Desnick & Schuchman, 2012). In phenylketonuria (PKU), the specific PAH variant genotype predicts residual phenylalanine hydroxylase activity, which in turn predicts responsiveness to sapropterin dihydrochloride (a pharmacological chaperone/cofactor for BH4-responsive PKU variants) before costly and time-consuming therapeutic trials are undertaken (Blau, Shen, & Carducci, 2014).

3.3 The Role of Modifier Genes and Polygenic Background

Even within the framework of Mendelian rare diseases, phenotypic variation among patients carrying identical primary disease-causing variants is often substantial, suggesting the influence of modifier genes and polygenic background on disease expression. These modifiers—which may include variants in genes encoding proteins that interact with, compensate for, or exacerbate the primary defect—can themselves have pharmacogenomic significance, influencing which patients respond to which therapies and to what degree. Understanding the modifier landscape of a given rare disease is therefore an increasingly important aspect of comprehensive pharmacogenomic characterization. In cystic fibrosis, for example, even among patients homozygous for the F508del CFTR variant, considerable variability in lung function decline and treatment response has been documented, attributable in part to variants in modifier genes including TGF-β1, MBL2, and genes encoding mucociliary clearance proteins (Cystic Fibrosis Foundation, 2023). Similarly, in sickle cell disease—a rare disease in the classical sense for certain ethnic groups—fetal hemoglobin (HbF) levels, a major modifier of disease severity, are themselves genetically determined by variants at the BCL11A locus and the HBS1L-MYB intergenic region, with implications for the response to hydroxyurea therapy (Luzzatto, Ally, & Notaro, 2021).

3. Disease-Specific Pharmacogenomic Applications

4.1 Inborn Errors of Metabolism

4.1.1 Phenylketonuria

Phenylketonuria represents one of the earliest and most instructive examples of pharmacogenomic precision in rare disease management. PKU results from pathogenic variants in the PAH gene encoding phenylalanine hydroxylase (PAH), the hepatic enzyme responsible for converting phenylalanine to tyrosine. With over 1,000 pathogenic PAH variants documented, PKU exemplifies the allelic heterogeneity characteristic of rare monogenic disorders. The majority of patients with PKU require lifelong adherence to a phenylalanine-restricted diet to prevent neurotoxicity from phenylalanine accumulation; however, approximately 20 to 50 percent of patients with mild-to-moderate hyperphenylalaninemia are responsive to treatment with sapropterin dihydrochloride, a synthetic form of the natural PAH cofactor tetrahydrobiopterin (BH4) (Blau, Shen, & Carducci, 2014). Sapropterin responsiveness is pharmacogenomically determined: patients carrying at least one allele associated with residual PAH activity—most commonly missense variants that disrupt folding rather than catalytic residues—are more likely to respond to BH4 cofactor supplementation, which stabilizes the residual enzyme and enhances its catalytic efficiency. Conversely, patients homozygous for null alleles (large deletions, frameshift mutations, or nonsense variants) have essentially no residual enzyme on which sapropterin can act and do not benefit from the drug.

4.1.2 Lysosomal Storage Disorders

The lysosomal storage disorders (LSDs) are a family of over 50 distinct conditions caused by deficient activity of lysosomal hydrolases, accessory proteins, or membrane transporters, resulting in the pathological accumulation of undegraded macromolecular substrates within lysosomes. Collectively affecting approximately 1 in 5,000 live births, the LSDs include Gaucher disease, Fabry disease, Pompe disease, Niemann-Pick disease, and many others, each with distinctive clinical presentations and molecular etiologies (Desnick & Schuchman, 2012). Enzyme replacement therapy (ERT) has been developed for a growing number of LSDs and represents a landmark achievement in the pharmacological management of inherited metabolic disorders. However, the efficacy of ERT is influenced by pharmacogenomic factors at multiple levels. First, the specific genotype of the patient—particularly whether pathogenic variants are null or missense—influences the severity of enzyme deficiency and the magnitude of clinical response to enzyme supplementation. Second, the immunogenicity of the administered recombinant enzyme varies with genotype: patients with complete absence of endogenous enzyme protein (null alleles) are at higher risk of developing high-titer anti-drug antibodies (ADAs) that neutralize ERT efficacy, compared with patients who produce some residual (albeit non-functional) enzyme protein and therefore have better immunological tolerance to the exogenous protein (Desnick & Schuchman, 2012). Third, variants in genes encoding lysosomal membrane proteins and accessory factors—including the cation-independent mannose 6-phosphate receptor through which ERTs are taken up by cells—may modify the tissue distribution and cellular uptake of administered enzyme, affecting clinical outcomes.

4.2 Cystic Fibrosis: The CFTR Modulator Paradigm

Cystic fibrosis stands as arguably the most mature and instructive example of pharmacogenomics-driven precision medicine in the entire rare disease landscape. The CFTR gene encodes the cystic fibrosis transmembrane conductance regulator, a chloride channel expressed in epithelial cells of the lungs, pancreas, intestines, and reproductive tract. Pathogenic CFTR variants are conventionally classified into six functional classes based on their molecular consequences: class I variants result in no protein synthesis; class II variants (including F508del, the most common) cause defective protein processing and trafficking; class III variants impair channel gating; class IV variants reduce channel conductance; class V variants reduce the quantity of normal CFTR; and class VI variants accelerate CFTR degradation (Cystic Fibrosis Foundation, 2023). This functional classification is directly pharmacogenomically actionable. CFTR potentiators (ivacaftor) increase the open probability of CFTR channels already present at the cell surface and are therefore effective primarily for class III gating mutations. This CFTR modulator paradigm—in which drug selection is entirely determined by the patient's specific genotype—has become a model for precision pharmacotherapy in rare diseases.

4.3 Rare Hematological Disorders

4.3.1 Sickle Cell Disease and Thalassemia

Sickle cell disease (SCD), caused by a single missense variant in the HBB gene (p.Glu6Val), illustrates how even a molecularly homogeneous disease generates pharmacogenomically relevant variability through genetic modifiers. The severity of SCD varies enormously among patients with identical primary HBB genotypes, largely due to differences in fetal hemoglobin (HbF) levels. HbF inhibits hemoglobin S polymerization, and patients with naturally elevated HbF have substantially milder disease. Hydroxyurea, the cornerstone pharmacological treatment for SCD, acts primarily by reactivating fetal hemoglobin production through epigenetic mechanisms. The magnitude of HbF response to hydroxyurea is itself partially heritable, with variants at the BCL11A locus and other HbF-regulatory loci predicting the degree of HbF induction and therefore the likely clinical benefit of treatment (Luzzatto, Ally, & Notaro, 2021).

4.3.2 Rare Coagulopathies

Von Willebrand disease (VWD), the most common inherited bleeding disorder, exemplifies pharmacogenomic diversity within a single rare disease category. VWD encompasses three main subtypes (types 1, 2, and 3) defined by quantitative or qualitative deficiency of von Willebrand factor (VWF), each with distinct genotypic underpinnings and distinct therapeutic requirements. Type 1 VWD, characterized by partial quantitative VWF deficiency, is often caused by heterozygous VWF missense variants and responds well to desmopressin (DDAVP), which releases VWF from endothelial storage sites. Types 2A, 2B, 2M, and 2N, caused by qualitative VWF defects, differ fundamentally in their response to DDAVP: type 2B VWD represents a contraindication to DDAVP because VWF release worsens thrombocytopenia due to enhanced platelet binding by the dysfunctional multimers. Type 3 VWD, resulting from near-complete VWF absence, requires exogenous VWF replacement and does not respond to DDAVP. Pre-treatment genotyping therefore enables clinicians to predict DDAVP response and avoid potentially hazardous therapy with inappropriate agents (Nicolosi, Baronciani, & Federici, 2020).

4.4 Rare Genetic Cardiomyopathies

Hereditary cardiomyopathies—including hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), and arrhythmogenic right ventricular cardiomyopathy (ARVC)—are individually rare but collectively significant causes of sudden cardiac death and heart failure. These conditions arise from pathogenic variants in genes encoding sarcomeric proteins (MYBPC3, MYH7, TNNT2), cytoskeletal proteins (LMNA, DES), or ion channel subunits (SCN5A, KCNQ1), each conferring distinct structural and electrophysiological phenotypes with different pharmacogenomic implications (Golbus et al., 2014).

4.5 Rare Neurological and Neuromuscular Disorders

The rare neurological and neuromuscular diseases constitute a large and heterogeneous category encompassing conditions such as spinal muscular atrophy (SMA), Duchenne muscular dystrophy (DMD), Rett syndrome, tuberous sclerosis complex (TSC), and many others. Pharmacogenomics intersects with these conditions at multiple levels, including variant-specific drug selection, pharmacokinetic modulation of drug metabolism, and the pharmacogenomic determinants of neuroinflammatory and neuroprotective responses. Spinal muscular atrophy provides a compelling example of mutation-specific therapeutic stratification. SMA results from homozygous deletion or mutation of the SMN1 gene; the severity of disease is modulated by the copy number of the paralogous SMN2 gene, which produces a small amount of functional SMN protein. Three distinct therapeutic modalities—nusinersen (an antisense oligonucleotide promoting SMN2 exon 7 inclusion), onasemnogene abeparvovec (an AAV9-delivered SMN1 gene therapy), and risdiplam (a small-molecule SMN2 splicing modifier)—each operate through different mechanisms and carry different practical considerations. Importantly, pre-symptomatic treatment enabled by newborn screening for SMN1 deletion is most effective in infants with higher SMN2 copy numbers, because residual SMN protein from multiple SMN2 copies confers greater neuronal reserve that can be leveraged by therapeutic augmentation (Friedman et al., 2017).

4. Technological Platforms Enabling Pharmacogenomic Discovery in Rare Diseases

5.1 Next-Generation Sequencing: Diagnostic Power and Pharmacogenomic Yield

Next-generation sequencing (NGS) technologies have fundamentally transformed the identification of disease-causing variants in rare disease patients and, by extension, enabled the pharmacogenomic characterization necessary for precision therapeutic decisions. Whole-exome sequencing (WES), which interrogates the protein-coding regions of the genome comprising approximately 1 to 2 percent of total genomic sequence, has achieved diagnostic yields of 25 to 50 percent in heterogeneous rare disease cohorts, with higher yields in selected patient populations defined by phenotypic severity or consanguinity (Biesecker & Green, 2014; de Ligt et al., 2012). Whole-genome sequencing (WGS) extends the diagnostic scope to non-coding regulatory and intronic regions, achieving even higher diagnostic yields in conditions where WES has been non-diagnostic, and simultaneously generating complete pharmacogenomic profiles of drug-metabolizing enzymes and drug targets from a single testing event (Lionel et al., 2018).

5.2 Bioinformatics and Variant Interpretation Challenges

The analytical pipeline from raw sequencing data to clinically actionable pharmacogenomic interpretation involves multiple complex bioinformatic steps, each introducing potential sources of error. Variant calling algorithms must accurately identify single-nucleotide variants, small insertions and deletions, copy number variants, and structural rearrangements against the backdrop of sequencing noise and alignment artifacts. Variant annotation must correctly classify variants according to population frequency, predicted functional impact, evolutionary conservation, and accumulated evidence from clinical databases. In the CYP2D6 gene (Chakravorty & Hegde, 2017). Variant classification frameworks, including those developed by the ACMG/AMP and by specialized disease expert panels, provide structured rubrics for assigning variants to pathogenicity categories (pathogenic, likely pathogenic, variant of uncertain significance, likely benign, benign). Variants of uncertain significance (VUS) are a particular challenge in pharmacogenomics: a VUS in CYP2C9 or another metabolizing enzyme may have real but uncharacterized functional consequences that can neither be ignored nor definitively acted upon clinically (Landrum et al., 2014).

5.3 Pharmacogenomic Databases and Clinical Decision Support

The translation of pharmacogenomic knowledge into clinical practice depends critically on the availability of curated, evidence-graded databases that summarize gene-drug relationships and provide actionable clinical recommendations. The PharmGKB knowledge base, the CPIC guidelines repository, and the DPWG database collectively represent the most comprehensive and regularly updated resources for clinical pharmacogenomics (Caudle et al., 2014; Swen et al., 2011). PharmGKB annotates variant-drug associations with levels of clinical annotation ranging from 1A (actionable, high evidence, with CPIC or DPWG guideline) through 4 (preliminary evidence, not clinically actionable), enabling clinicians and informaticians to prioritize actionable pharmacogenomic findings in the context of clinical decision support. The integration of pharmacogenomic decision support into electronic health record (EHR) systems is a critical step toward routine clinical implementation. Pioneering institutions including St. Jude Children's Research Hospital, Vanderbilt University Medical Center, and several European academic medical centers have implemented pre-emptive genotyping programs in which patients' pharmacogenomic profiles are determined once and stored in the EHR, generating automated prescribing alerts whenever a clinician orders a medication affected by the patient's genotype (Hicks et al., 2013; Manolio et al., 2013).

5.4 Newborn Screening and Early Pharmacogenomic Identification

Newborn screening (NBS) programs represent a uniquely powerful vehicle for early pharmacogenomic identification in rare diseases. By identifying affected infants before symptom onset—typically through tandem mass spectrometry-based metabolite profiling supplemented by molecular confirmation—NBS enables the earliest possible initiation of disease-modifying therapy, during the developmental window in which treatment is most effective. Levy (1999) documented the transformative impact of expanded NBS by tandem mass spectrometry on early diagnosis of amino acid, organic acid, and fatty acid oxidation disorders, a category of inborn errors of metabolism in which prompt dietary or pharmacological intervention can prevent irreversible neurological damage.

The integration of genomic sequencing into NBS programs—as piloted in programs such as BabySeq and several national genomic medicine initiatives—offers the prospect of simultaneously establishing molecular diagnosis and characterizing secondary pharmacogenomic variants at birth, creating comprehensive genomic profiles that can guide therapeutic decision-making from the earliest stages of life (Friedman et al., 2017). This approach raises important considerations about consent, data storage, secondary findings, and the psychological impact of genomic information on families, which are addressed further in the section on ethical considerations.

5. Clinical Implementation of Pharmacogenomics in Rare Disease Management

6.1 Pre-Emptive versus Reactive Genotyping Strategies

Two broad strategic approaches to clinical pharmacogenomic implementation have been advocated: reactive genotyping, in which genetic testing is performed in response to a specific clinical need (e.g., before initiating a drug with a known pharmacogenomic risk), and pre-emptive genotyping, in which a comprehensive pharmacogenomic profile is established at a single point in time (typically at first clinical encounter or upon diagnosis) and stored for application to all future prescribing decisions. For rare disease patients, the pre-emptive approach has compelling advantages: these patients typically initiate complex therapeutic regimens early in life, receive multiple drugs over many years, and often have limited clinical monitoring capacity due to geographic isolation from specialized centers (Manolio et al., 2013). Pre-emptive genotyping panels in rare disease settings should encompass at minimum the pharmacogenomic loci endorsed by CPIC and DPWG for clinically actionable drug-gene pairs, including CYP2D6, CYP2C19, CYP2C9, CYP3A5, DPYD, TPMT/NUDT15, SLCO1B1, UGT1A1, and G6PD. When WES or WGS has been performed for diagnostic purposes—as is increasingly the case for rare disease patients—these pharmacogenomic profiles can be bioinformatically extracted from existing data without additional laboratory cost, maximizing the value of already-acquired genomic information (Aronson & Rehm, 2015).

6.2 Multidisciplinary Team Approaches

Effective clinical implementation of pharmacogenomics in rare disease management requires a multidisciplinary team model that integrates the expertise of clinical geneticists, clinical pharmacologists, genetic counselors, pharmacy specialists, disease-specific specialists (e.g., metabolic dietitians, pulmonologists, hematologists), and informatics professionals. No single specialty possesses the full breadth of knowledge required to translate pharmacogenomic test results into optimal therapeutic decisions in the complex context of rare disease management. The rarity and complexity of these conditions, combined with the rapidly evolving evidence base, makes ongoing multidisciplinary collaboration essential rather than optional (Kaufmann, Pariser, & Austin, 2018). The genetic counselor plays a particularly important role in the pharmacogenomic component of rare disease management. Genetic counselors are uniquely trained to communicate complex genomic information to patients and families, to contextualize the clinical significance of pharmacogenomic variants within the broader diagnostic picture, to obtain informed consent for genomic testing, and to provide psychosocial support in the context of genomic uncertainty. As pharmacogenomic testing becomes more routine in rare disease care, the integration of genetic counselors into medication management teams—rather than confining their role to diagnostic counseling—will become increasingly important (Scott, 2011).

6.3 Challenges in Clinical Trials for Rare Disease Pharmacogenomics

The design and conduct of pharmacogenomically stratified clinical trials in rare diseases faces formidable practical challenges. Patient populations are inherently small, making randomized controlled trials with adequate statistical power extremely difficult without multi-institutional or multinational collaboration. The high allelic heterogeneity of most rare diseases means that even within a single disease, the patient population is often genotypically fragmented into multiple small subgroups with potentially distinct pharmacogenomic profiles, each of which would ideally require separate evaluation. And the ethical imperative to provide effective treatment—particularly for life-threatening conditions—can make placebo-controlled trial designs ethically problematic (Sawicki & Celinska-Lowenhoff, 2021). Innovative trial designs have been developed to address these challenges. Adaptive clinical trial designs allow pre-specified modifications to sample size, treatment arms, or eligibility criteria based on accumulating data, maximizing statistical power while minimizing patient exposure to inferior treatments. Basket trials enroll patients based on molecular characteristics—such as the presence of a specific mutation class—across multiple disease contexts, enabling pharmacogenomic hypotheses to be tested across a broader patient population than any single rare disease would permit. Natural history studies and patient registries, increasingly incorporating genomic data, provide valuable reference datasets that enable historical control comparisons and Bayesian adaptive designs (IRDiRC, 2022).

6.4 Regulatory Frameworks Supporting Pharmacogenomic Development

Regulatory agencies in major markets have developed specific frameworks to encourage the integration of pharmacogenomics into drug development and approval. The FDA's table of pharmacogenomic biomarkers in drug labeling—currently encompassing over 400 drug-biomarker pairs—provides a regulatory mandate for pharmacogenomic testing when evidence warrants, and creates a framework for drug developers to incorporate pharmacogenomic eligibility criteria and dose modification recommendations into approved labeling (European Medicines Agency, 2022). The orphan drug designation programs of the FDA (Orphan Drug Act, 1983) and EMA (Regulation (EC) No 141/2000) provide financial incentives—including seven to ten years of market exclusivity, fee reductions, and enhanced regulatory assistance—that partially offset the commercial challenges of developing drugs for small patient populations. FDA's Breakthrough Therapy Designation, Accelerated Approval, and Priority Review pathways have collectively enabled more rapid approval of pharmacogenomically targeted therapies for rare diseases, allowing earlier patient access while post-marketing confirmatory studies are completed. The EMA's PRIME (PRIority MEdicines) scheme serves a similar purpose in the European regulatory context, providing enhanced scientific and regulatory support to developers of medicines that target unmet medical needs, including rare diseases with pharmacogenomically defined patient populations (European Medicines Agency, 2022).

6. Summary of Key Pharmacogenomic Associations in Selected Rare Diseases

Table 1. Key pharmacogenomic gene-drug associations in selected rare diseases. ADA = anti-drug antibody; ERT = enzyme replacement therapy; HbF = fetal hemoglobin; HCM = hypertrophic cardiomyopathy; PM = poor metabolizer; VWF = von Willebrand factor