Multiple sclerosis (MS) is a clinically and biologically heterogeneous, immune-mediated disease of the central nervous system, with substantial interindividual variability in disease course and response to disease-modifying therapies (DMTs). Over the past three decades, the MS therapeutic landscape has expanded considerably; however, treatment selection and switching remain guided primarily by clinical phenotype and imaging findings rather than molecular predictors of response. Despite extensive clinical trial evidence, prospectively identifying responders and non-responders to specific DMTs remains challenging.
Genetic variability appears to influence differences in treatment efficacy, tolerability, and long-term outcomes in people with MS. Numerous candidate pharmacogenomic variants have been reported across interferon-β, glatiramer acetate, oral agents, and monoclonal antibodies; nevertheless, replication has been inconsistent, effect sizes are modest, and no genetic marker has yet been clinically validated for routine use. Consequently, pharmacogenomics is largely absent from current MS treatment algorithms.
This review critically evaluates the existing pharmacogenomic literature across approved DMTs, highlighting reproducible findings, methodological limitations, and gaps that hinder clinical translation. We further discuss requirements for integrating pharmacogenomic markers into routine practice, emphasizing the need for large, multiethnic cohorts, standardized response definitions, and functional validation. Overall, these insights underscore both the potential and current limitations of pharmacogenomics in advancing precision medicine for MS.
INTRODUCTION
Multiple sclerosis (MS) is a complex, immune-mediated disorder with a multifactorial etiology shaped by the interplay of genetic predisposition, epigenetic regulation, and environmental exposures.1 The treatment of MS progresses according to disease manifestations and involves several stages: management of severe attacks, administration of disease-modifying therapies (DMTs) that reduce MS biological activity, and symptom-targeted interventions.2 Integrating clinical phenotypes and magnetic resonance imaging (MRI) characteristics with demographic data enables a precision-based approach to MS management, guiding transitions between acute attack therapy, disease-modifying protocols, and symptomatic support.3 Currently, no treatment completely eradicates MS. Predicting long-term prognosis and selecting the most appropriate therapeutic approach at diagnosis in people with MS (pwMS) are crucial for refining personalized treatment strategies.
The heterogeneity of MS necessitates a departure from a one-size-fits-all approach, as treatment benefits vary significantly across the patient population.4 This multifactorial nature complicates the identification of universal, reliable biomarkers, making personalized prognostic tools essential for modern clinical practice. Variability in response to DMTs in pwMS presents a significant clinical challenge, as delays in identifying effective therapies may expose patients to adverse effects without substantial benefit. Multi-parametric machine learning frameworks that integrate polygenic risk scores with clinical, imaging, and laboratory variables have shown promise in predicting treatment responses in MS; however, despite improved classification accuracy, no single reproducible genetic variant has yet emerged for routine clinical implementation.5
Ongoing research aims to validate potential pharmacogenomic variants in large, well-characterized patient populations, yet the routine use of pharmacogenetics in MS remains a distant goal. While genomic variation is a key determinant of this variability, identifying reliable pharmacogenomic markers is often hindered by the lack of standardized clinical definitions and the oversight of confounding variables. As illustrated in Figure 1, evaluating the multifactorial determinants of MS pharmacogenomics is essential to distinguish true biological drug non-responders. This framework requires the integration of demographic, clinical baseline, and environmental factors alongside high-resolution genomic data to capture the full landscape of treatment outcomes. We addressed this complexity by utilizing whole-exome sequencing (WES) and drug-associated HLA allele typing within a Turkish Familial MS (TuFaMS) cohort to identify specific risk alleles, including DQB106:02 and DQA101:02, which are significantly associated with responses to Fingolimod and Ocrelizumab. By coupling these findings with pathway enrichment analysis, we identified significant involvement of cytokine signaling pathways, further validated by differential gene expression in Fingolimod responders and non-responders from RNA-seq data (GSE250453).6 This integrated approach demonstrates that combining clinical metrics with high-resolution genomic profiling can bridge the gap between statistical associations and a functional, mechanistic understanding of individual treatment trajectories in MS. By consolidating evidence across multiple pharmacogenomic studies, this review characterizes the collective impact of genetic variation on the heterogeneity of DMT therapeutic responses in MS.
Clinical phenotyping and covariates in MS pharmacogenomics
Effective pharmacogenomic analysis in MS requires a clear distinction between biological treatment failure and alternative reasons for medication switching. To ensure the accuracy of genetic associations, defining treatment non-response is critical; the core phenotype should be characterized by breakthrough disease activity despite adherence, including new clinical relapses, confirmed disability progression (Expanded Disability Status scale increase), or new/enlarging T2 or Gadolinium-enhancing lesions on MRI. Confounding Discontinuation: Medication switches due to allergic reactions, adverse events (e.g., lymphopenia, liver enzyme elevation), or patient non-adherence must be treated as confounders and distinguished from pharmacodynamic non-response.
Demographic covariates such as age at onset, biological sex, and ancestry are essential because they independently influence both immune baseline and therapeutic trajectories. Environmental modulators, including vitamin D levels, smoking status, and BMI, should be incorporated into multifactorial models due to their established role in influencing drug efficacy.
DMTs in MS
Therapeutic management of MS encompasses a broad spectrum of DMTs, including immunomodulators, immunosuppressants, and targeted monoclonal antibodies. These consist of injectable formulations such as subcutaneous or intramuscular interferons (IFNs) and subcutaneous glatiramer acetate (GA); intravenous natalizumab, mitoxantrone, alemtuzumab, and ocrelizumab; and orally administered agents including fingolimod, teriflunomide, and dimethyl fumarate (DMF). Collectively, these therapies have demonstrated robust efficacy in reducing annualized relapse rates during the initial phases of the disease.7
According to the European Medicines Agency, the initial era of DMT development in the 1990s established the foundation with injectable agents such as IFN beta (IFN-β)-1b (Betaferon®, 1995), IFN-β-1a (Avonex®, 1997, and Rebif®, 1998), and Mitoxantrone (Novantrone®, 1998). Subsequent advancements in the 2000s introduced GA (Copaxone®, 2004) and the targeted monoclonal antibody Natalizumab (Tysabri®, 2006) (Figure 2). Between 2010 and 2015, orally bioavailable agents such as Fingolimod (Gilenya®, 2011) and Teriflunomide (Aubagio®, 2013) emerged alongside potent intravenous treatments, including Alemtuzumab (Lemtrada®, 2013). The most recent approvals, including Ocrelizumab (Ocrevus®, 2018) and the S1P modulators Siponimod (Mayzent®, 2020), Ponesimod (Ponvory®, 2021), and Ofatumumab (Kesimpta®), further reflect the growing heterogeneity of therapeutic options (Figure 2).
Despite this diversity, IFN-β and GA remain the most commonly prescribed first-line interventions worldwide (Figure 3).8 These agents offer substantial clinical benefits, including reduced relapse risk, slower disability progression, and improved MRI parameters, while maintaining a relatively mild adverse effect profile. Nevertheless, their therapeutic effect is incomplete, and the magnitude of benefit varies widely among individuals. Current evidence suggests that approximately 30–50% of patients fail to meet established response benchmarks, a variability thought to be driven in part by interindividual genetic differences.9,10 In clinical practice, therapeutic decision-making in MS typically follows a line-of-therapy paradigm, stratifying DMTs into first-, second-, and third-line options according to relative efficacy, safety profile, and cumulative risk (Figure 3).11 First-line therapies generally provide moderate efficacy with favorable long-term tolerability, whereas second- and third-line agents include high-efficacy therapies associated with more pronounced immunosuppressive effects and monitoring requirements. Although this hierarchical framework facilitates standardized treatment algorithms, it assumes a stepwise and homogeneous disease trajectory, inadequately accounting for interindividual variability in disease activity, treatment response, and long-term prognosis.12
Given the heterogeneity of MS pathogenesis and treatment response, identifying genomic variants that predict individual responsiveness to specific DMTs is crucial.13 Pharmacogenomic profiling represents a promising strategy for patient stratification, enabling the identification of molecular signatures that correlate with differential therapeutic efficacy and tolerability.14,15 Early identification of likely responders could facilitate timely initiation of the most appropriate DMT, maximizing disease control, reducing the risk of irreversible neuroaxonal loss, and minimizing exposure to ineffective or poorly tolerated therapies. Integrating pharmacogenomic marker-driven strategies into routine clinical pathways would support a transition from empirical, trial-and-error prescribing to a precision medicine model, in which treatment selection is guided by objective, patient-specific molecular and clinical data.
Selection criteria and literature search strategy
To ensure a comprehensive and systematic overview of the pharmacogenomic landscape in MS, a two-tiered search strategy was implemented. First, the PharmGKB (Pharmacogenomics Knowledgebase) database was queried as of January 2026 to identify genetic variants with established clinical annotations and evidence levels using standardized MeSH terms and keywords, including “Multiple Sclerosis.” It is important to note that all identified genetic variants associated with MS DMTs currently hold Level 3 evidence within the database. This evidence level signifies suggestive clinical associations that have not yet met the high evidentiary thresholds required for inclusion in formal clinical guidelines or Food and Drug Administration drug labels.
This was followed by a systematic literature search in PubMed/MEDLINE using a combination of MeSH terms and keywords: (“Multiple Sclerosis” OR “MS”) AND (“Pharmacogenomics” OR “Pharmacogenetics”) AND (“Genetic Variant” OR “SNP”) AND (“Treatment Response” OR “Drug Toxicity”) AND (“Disease-Modifying Therapies” OR “DMTs” OR “Dimethyl Fumarate” OR “Fingolimod” OR “Glatiramer Acetate” OR “Interferon-beta” OR “Natalizumab” OR “Ocrelizumab” OR “Teriflunomide”).
The selection process adhered to strict inclusion and exclusion criteria to maintain translational relevance. Inclusion criteria encompassed: (i) clinical studies involving human subjects reporting statistically significant associations between specific genotypes and outcomes, such as relapse rate, EDSS progression, or MRI activity; and (ii) functional studies using validated human-derived cell lines (e.g., Jurkat, T cells, or B cells) to investigate molecular mechanisms, including gene expression or signaling pathways under drug exposure. Exclusion criteria applied to: (i) studies relying exclusively on non-human animal models (e.g., EAE); (ii) research focusing solely on disease susceptibility without assessing treatment response; and (iii) low-quality data lacking sufficient statistical power, missing p-values, or non-peer-reviewed sources, such as conference abstracts.
By synthesizing curated database evidence with mechanistic insights from cell-based assays, this review prioritizes variants supported by both clinical relevance and functional validation.
Genetic variants associated with treatment response in MS
IFN-β is a first-line therapy available as intramuscular or subcutaneous formulations of recombinant IFN-β-1a and subcutaneous IFN-β-1b. Its mechanism of action involves binding to type I IFN receptors (IFNAR1/2) and activating the JAK–TYK2–STAT signaling pathway, which shifts the cytokine profile toward an anti-inflammatory state, decreases antigen presentation, and limits T-cell migration into the central nervous system (CNS).16 Pharmacogenomic studies have identified variants in IFNAR1, IFNAR2, TYK2, and downstream IFN-stimulated genes such as OAS1 and MX1 as potential modulators of therapeutic response.16,17 Transcriptomic “type I interferon signature” gene expression in monocytes has been associated with poorer response to IFN-β in MS.18 A recent genome-wide association analysis reported that rs7665090, located near NF-κB–related genes, is associated with favorable clinical outcomes under IFN-β therapy (Supplementary Table),19 whereas earlier studies suggested that common IFNAR1/IFNAR2 variants may influence MS susceptibility but do not robustly predict IFN-β response.17 Certain HLA class II alleles (HLA-DRB10401 and HLA-DRB10408) and innate immune pathway variants have been linked to an increased risk of neutralizing antibody formation, which can reduce drug efficacy.20,21 In IFN-β–treated relapsing-remitting MS (RRMS), pharmacogenomic signals have been reported at both single-locus and polygenic levels. Replication data support an association between GPC5 and clinical response, most notably rs10492503 (p = 0.0005), while HAPLN1 variants showed no association in the same analysis.22 Complementing this, a genome-wide pharmacogenomic study of 206 patients identified significant genotype-frequency differences between responders and non-responders at multiple loci, including GPC5, COL25A1, HAPLN1, CAST, and NPAS3.23 Candidate-gene analyses demonstrated that allelic combinations were informative, with JAK2–IL10RB–GBP1–PIAS1 (permutation p = 0.0008) and JAK2–IL10–CASP3 (p = 0.001) differing in frequency between response groups.24 Positive response to IFN-β treatment was significantly associated with MXA rs464138 AA, MXA rs2071430 G, and MXA rs17000900 GG genotypes (p < 0.0001, 0.015, and 0.018, respectively) (Supplementary Table 1).25-27
Within the IFNAR1 gene, the rs1012335 G allele has been linked to a negative response to IFN-β therapy (p = 0.036),27,28 and IRF5 rs2004640 (p = 0.0006) has been associated with poor pharmacological response, characterized by increased T2 lesion burden (Supplementary Table 1).29 The rs4774388 variant within the RORA gene has also been identified as significantly associated with reduced responsiveness to IFN-β therapy in patients with MS.30
Additional genetic variations associated with suboptimal or negative responses to IFN-β in RRMS include CD46 (rs2724385),31 CD58 (rs12044852),32 GAPVD1 (rs10819043, rs10760397),25 GPC5 (rs10492503, rs1411751), GABRB3 (rs832032),33 IRF5 (rs2004640), MXA (rs464138), PELI3 (rs2277302),33 and ZNF697 (rs10494227). Conversely, GAPVD1 (rs2291858)25 and FHIT (rs760316)25 have been linked to positive responses to IFN-β therapy (Supplementary Table 1).34
Glatiramer acetate
GA, a first-line therapy administered subcutaneously, is a synthetic myelin-mimetic copolymer that promotes a shift in T-cell polarization from proinflammatory Th1 toward anti-inflammatory Th2 and regulatory T-cell phenotypes while also modulating antigen-presenting cell function.35,36 The most consistent pharmacogenomic signals for GA reside within the HLA region rather than individual non-HLA SNPs. Multiple cohort studies have linked HLA-DRB1*15:01 and related class II haplotypes with differential therapeutic responses, showing in some populations better relapse control and in others poorer outcomes, reflecting substantial population-specific heterogeneity and limited replication.37 HLA association analyses indicate that the presence of DR15 or DQ6, or the absence of DR17 and DQ2 alleles, is associated with favorable clinical response. Specifically, the DR15-DQ6 positive/DR17-DQ2 negative haplotype combination strongly predicts a good response (71%), whereas the DR15-DQ6 negative/DR17-DQ2 positive combination strongly predicts a poor response (17%) (Supplementary Table 1).37 In a cohort of 139 MS patients stratified as GA responders (n = 81) and non-responders (n = 58), genotyping of HLA-DRA (rs3135388, rs3135391), HLA-DQA1 (rs9272346), and IL6 (rs1800795, rs1900796) showed no significant association with treatment response, with female predominance observed in both groups. Non-responders exhibited higher EDSS scores and greater MRI lesion burden, highlighting the need for larger, adequately powered pharmacogenetic studies to define robust GA response markers and enable personalized therapy.38
Among GA-treated MS patients, HLA-DRB115:01 was associated with decreased treatment response in a European cohort (p = 0.0056). The risk of non-response increased when DRB115:01 co-occurred with the rs1800469 A allele and further rose with combined haplotypes including the rs333 (CCR5) deletion and rs1012335 G allele.39 In the same study, IFNB1 rs1051922 (G vs. A) and CTLA4 rs231775 (G vs. A) showed no association with GA response. Beyond HLA, the EOMES rs2371108 T allele was linked to greater GA responsiveness in Europeans (p = 0.018), remaining significant after multiple-testing correction when responders were contrasted against non-responders plus intermediate responders.28 Overall, these findings suggest predominantly HLA-anchored effects with a few non-HLA candidate variants, alongside several null or inconsistent results, underscoring the need for larger, well-controlled pharmacogenomic studies prior to clinical implementation. Additionally, the rs1799752 polymorphism within the ACE gene has been associated with a negative response to IFN-beta therapy, showing a particularly notable correlation with treatment outcomes in male patients.40
Teriflunomide
Teriflunomide, an oral first-line therapy, is a selective, non-competitive inhibitor of dihydro-orotate dehydrogenase (DHODH), an essential mitochondrial enzyme for de novo pyrimidine synthesis. By limiting the clonal expansion of activated T and B lymphocytes while sparing resting cells, teriflunomide modulates immune responses in MS.41 Unlike GA or IFN-β, pharmacogenomic research on teriflunomide in MS is sparse, and no single-nucleotide polymorphism (SNP), either within HLA or non-HLA regions, has been reproducibly associated with therapeutic response in prospective cohorts. Candidate studies have examined DHODH gene variants, including rs3213422 and rs3213421, which in other autoimmune diseases (e.g., rheumatoid arthritis) have been linked to altered leflunomide or teriflunomide metabolism; however, these variants have not shown a clear responder/non-responder effect in MS.42 To date, no validated HLA or non-HLA SNP can be used to stratify teriflunomide responders from non-responders in clinical practice, and treatment selection continues to rely on clinical, MRI, and safety profiles rather than genetic testing.
Dimethyl fumarate
DMF, an oral first-line therapy, is rapidly hydrolyzed to its active metabolite, monomethyl fumarate, which penetrates immune cells and activates the NRF2 antioxidant pathway through covalent modification of KEAP1 cysteine residues.43,44 In MS, NRF2 activation reduces oxidative stress–mediated axonal injury and shifts immune cell phenotypes toward an anti-inflammatory profile.45 DMF decreases the proportion of proinflammatory Th1 and Th17 lymphocytes, enhances regulatory T-cell function, and promotes neuroprotective phenotypes in microglia and astrocytes. It also inhibits NF-κB signaling in antigen-presenting cells, reducing the production of proinflammatory cytokines such as interleukin (IL)-1β, IL-6, and tumor necrosis factor-α, thereby limiting CNS infiltration by activated lymphocytes. A recent patient-level, population-specific pharmacogenomic study identified two promoter polymorphisms in the long intergenic non-coding RNA linc00513 with drug-specific effects across DMTs: rs205764 (G allele) was associated with a significantly lower response to DMF, whereas rs547311 showed no significant association with DMF response but correlated with higher EDSS.46
Fingolimod
Fingolimod (FTY720), an oral second-line therapy, is a sphingosine analog phosphorylated in vivo to its active phosphate form, which binds with high affinity to sphingosine-1-phosphate receptor subtype 1 (S1PR1) on lymphocytes.47 This interaction induces receptor internalization and functional antagonism, preventing lymphocyte egress from secondary lymphoid tissues into the circulation.48 In MS, this mechanism reduces CNS infiltration by autoreactive T and B cells, thereby lowering inflammatory lesion activity and relapse rates.49,50 Pharmacogenomic studies in MS have identified the long intergenic non-coding RNA promoter variant linc00513 rs205764 as a potential responder marker, associated with greater relapse rate reduction in fingolimod-treated patients, whereas rs547311 in the same locus showed no association with treatment response.46 Direct analyses of S1PR1 coding and regulatory SNPs in MS cohorts have not demonstrated significant associations with clinical efficacy or immunologic parameters such as IL-17 modulation, suggesting limited predictive value for single-locus S1PR1 variation.51 To date, despite consistent cohort-level associations for rs205764, no validated HLA or non-HLA SNP can be implemented in routine clinical practice to stratify fingolimod responders from non-responders in MS. A strong association between the MHC region and MS susceptibility has been identified, with HLA-DRB11501 having the most significant impact.52 Variants linked to DMTs in MS include HLA risk alleles DQA101:02, DQB106:02, DRB115:01, DQA103:01, and DQB103:02. An analysis of 138 WES datasets from the TuFaMS cohort assessed medication switches and genetic correlations.53 Fisher’s exact test indicated significant associations between drug-HLA allele pairs, with the strongest link observed between fingolimod and DQB1*06:02 (p = 0.0166), present in five individuals.53
Natalizumab
Natalizumab, a second-line intravenous therapy, is a humanized monoclonal antibody targeting the α4 subunit of VLA-4 (α4β1 integrin/CD49d) on lymphocytes. By blocking α4-integrin–VCAM-1 interactions, it prevents firm adhesion and transmigration of leukocytes across the blood–brain barrier (BBB), thereby suppressing CNS inflammatory activity in MS.54 The largest genome-wide association study investigating pharmacogenomic response to natalizumab in MS revealed no single variant reaching genome-wide significance. However, polymorphisms within the Wnt/β-catenin signaling pathway, which is critical for BBB formation and maintenance, have been identified as potential modulators of treatment response.55 These variants lead to downregulated β-catenin–mediated transcriptional activity, resulting in a leaky barrier phenotype that facilitates continuous leukocyte diapedesis and neuroinflammation, bypassing the mechanism of action of systemic DMTs. Consequently, these findings suggest that clinical non-responsiveness to natalizumab is significantly influenced by CNS-specific structural genetics, in which genetically determined failure of endothelial homeostatic repair limits the efficacy of current immunomodulatory protocols.55
Mitoxantrone
Mitoxantrone, a synthetic anthracenedione derivative, exerts cytotoxic effects by inhibiting topoisomerase II, impairing DNA repair and replication, and reducing lymphocyte and macrophage proliferation. This action decreases proinflammatory cytokine release and suppresses myelin degradation.56 Two studies have investigated the pharmacogenetic associations between mitoxantrone response and genetic polymorphisms, yielding conflicting results. The first study identified SNPs in ABC-transporter genes (ABCB1 and ABCG2) as potential pharmacogenetic markers associated with clinical response in patients with RRMS or secondary progressive MS (SPMS). In contrast, the second study, which included patients with primary progressive MS (PPMS), did not confirm any significant association, despite observing clinical response rates of 53.7% in PPMS and 78.1% in RR/SPMS (p = 0.039), with no correlation between treatment efficacy and ABCB1 or ABCG2 genotype.57,58
DISCUSSION
Several limitations hinder the translation of pharmacogenomic discoveries into clinical practice. Many studies are constrained by small sample sizes, limited replication, and a lack of diversity in patient populations. Differences in study design, treatment regimens, and outcome measures further complicate comparisons of results. Moreover, most findings originate from association studies without extensive functional validation, leaving the mechanistic underpinnings of many candidate variants unresolved.
A precise definition of treatment nonresponse is the cornerstone of pharmacogenomic research, as it must be strictly distinguished from voluntary discontinuation or allergic reactions. True therapeutic failure should be characterized by breakthrough disease activity despite treatment, specifically manifested as clinical relapses, disability progression, or increased MRI activity. Establishing a consensus on this definition is the first step toward a functional understanding of drug efficacy. Once defined, it is essential to integrate demographic, clinical, biological/environmental, and genomic factors, along with treatment history, into a unified analytical framework. This holistic approach ensures that genetic variants are evaluated not in isolation but as components of a complex biological system that dictates individual clinical trajectories. Ultimately, correlating these multilayered data points with functional assays in cell lines enables a transition from simple statistical associations to a comprehensive functional understanding of treatment response in MS.
From a clinical perspective, integrating pharmacogenomic insights into MS management could facilitate the development of more personalized treatment algorithms. Identifying genetic predictors of drug efficacy or toxicity may allow clinicians to optimize therapy selection, minimize unnecessary exposure to ineffective agents, and enhance patient outcomes. However, despite encouraging signals from candidate gene studies, no robust pharmacogenomic marker has yet reached clinical implementation for MS. Current therapeutic decisions remain largely guided by clinical features, imaging findings, and patient preferences rather than genetic data.
Although B-cell–depleting therapies currently represent the most potent pharmacological interventions in the MS landscape, the specific genomic variants modulating individual responses to these high-efficacy agents remain largely uncharacterized. Future pharmacogenomic research must focus on identifying the molecular determinants of treatment failure to elucidate why a subset of patients exhibits breakthrough disease activity despite profound peripheral B-cell lymphopenia. Establishing these genetic profiles will be foundational for shifting the therapeutic paradigm toward a precision-based model, ensuring optimized selection of anti-CD20 protocols based on an individual’s unique molecular signature. Moreover, comprehensive analyses of treatment outcomes, particularly in patients who require third-line therapies yet continue to show suboptimal responses, may facilitate the identification of novel variants associated with drug resistance or reduced efficacy, thereby expanding the scope of pharmacogenomic markers relevant to personalized treatment strategies in MS.
The PharmGKB database provides a curated collection of pharmacogenomic associations categorized by levels of evidence, ranging from preliminary findings (Level 4) to clinically actionable variants supported by replicated studies and clinical guidelines (Level 1A).59 These evidence levels standardize the strength of genotype–drug response relationships and guide translation into clinical practice. For MS, only a limited number of variants related to DMT response are listed in PharmGKB, most of which remain at lower evidence levels due to insufficient replication. According to PharmGKB classifications of variant–drug associations, as of January 2026, no clinically actionable variants with Level 1A or 1B evidence have been identified for DMTs in MS. This highlights the existing gap between research findings and clinically validated pharmacogenomic implementation in MS therapy.
Advancement in pharmacogenomics for MS relies on identifying genetic variations and specific antibody profiles to predict disease trajectory and personalize therapeutic interventions. Recent evidence indicates that functional polymorphisms within the SNARE complex, particularly the VAMP2 Del/Del genotype and the synaptotagmin XI C allele, are significantly associated with increased MS susceptibility.60 These genetic markers represent potential targets for novel medications designed to restore synaptic homeostasis and mitigate synaptopathy in early disease phases. Furthermore, the presence of anti-MOG-IgG serves as a critical biomarker influencing the selection of immunomodulatory treatments, as patients with this profile may require strategies such as rituximab or azathioprine to prevent relapses.61 Integrating pharmacogenomic data with prognostic indicators, including advanced age at onset and spinal cord involvement, is essential for optimizing clinical outcomes and achieving precision management in neuroimmunology.
Future research should prioritize large-scale, multiethnic cohorts and integrate genomic data with other molecular layers, such as transcriptomics, epigenomics, and proteomics. Functional studies using advanced tools, including CRISPR-based genome editing and in vitro immune cell models, are critical to validating the biological relevance of identified variants. Ultimately, translating pharmacogenomic findings into clinical guidelines will require concerted efforts across genetics, neurology, and pharmacology. Establishing robust, clinically validated biomarkers could enable the integration of pharmacogenomics into personalized medicine for MS, reducing trial-and-error prescribing and improving long-term therapeutic outcomes.
CONCLUSION
This review highlights the substantial effort invested in identifying pharmacogenomic determinants of treatment response in MS and underscores the complexity of translating these findings into clinical practice. While numerous genetic variants have been associated with differential responses to DMTs, most reported associations originate from small or population-specific cohorts, lack robust replication, and are rarely supported by functional validation. Consequently, no pharmacogenomic marker has yet met the evidentiary threshold required for incorporation into clinical guidelines for MS.
Nevertheless, the collective evidence supports the role of genetic variability in shaping interindividual differences in treatment efficacy and disease trajectory. The challenge moving forward is not the absence of candidate signals but the need to distinguish clinically meaningful predictors from spurious associations. Addressing this challenge will require large, prospective, multiethnic studies with harmonized outcome measures, integration of genomic data with transcriptomic and immunophenotypic layers, and mechanistic validation using experimental models.
Ultimately, the successful integration of pharmacogenomics into MS care has the potential to shift therapeutic decision-making away from trial-and-error approaches toward a precision-based framework, enabling earlier optimization of therapy and improved long-term outcomes. Until such markers are rigorously validated, pharmacogenomic findings should be interpreted cautiously and regarded as hypothesis-generating rather than actionable tools.
Authorship Contributions: Concept- M.D.R., A.S., E.T.T.; Design- M.D.R., E.T.T.; Supervision- A.S., E.T.T.; Data Collection or Processing- M.D.R.; Analysis and/or Interpretation- M.D.R.; Literature Review- M.D.R.; Writing- M.D.R., E.T.T.; Critical Review- A.S., E.T.T.
Conflict of Interest: The other authors declared no conflicts of interest.
Funding: The authors declared that this study received no financial support.
Peer-Review: Eda Tahir Turanlı is a member of the Editorial Board of the Balkan Medical Journal. However, she was not involved in the editorial decision of the manuscript at any stage.
Supplementary Table 1: http://balkanmedicaljournal.org/img/files/Table-1.xlsx
Copyright © 2026 - Balkan Medical Journal
){kind=link}
){kind=link}
){kind=link}