Glycomic Insights in Gynecological Disease: From Molecular Mechanisms to Precision Diagnostics and Therapeutics

Glycosylation in eukaryotic cells proceeds through two primary pathways that differ in initiation site and subcellular localization [ 13 ]. N-linked glycosylation initiates in the endoplasmic reticulum with transfer of a preformed high-mannose glycan to asparagine residues within consensus sequences (Asn-X-Ser/Thr), followed by sequential processing in the Golgi apparatus to generate complex-type N-glycans with diverse branching architectures [ 14 ]. The biosynthetic precision of N-glycosylation is governed by expression levels and catalytic activities of approximately 100 distinct N-glycan processing glycosyltransferases and glycosidases, creating a delicate equilibrium that distinguishes healthy from pathological states [ 15 ]. O-linked glycosylation, encompassing mucin-type modifications where glycans attach to serine and threonine residues, begins with transfer of N-acetylgalactosamine (GalNAc) by peptidyl-N-acetylgalactosaminyltransferases (GALNTs) in the Golgi apparatus, followed by optional extension through additional enzymatic steps [ 16 ]. In healthy cells, O-glycans undergo orderly maturation through coordinated enzymatic action; in pathological states, this process is characteristically interrupted at early steps, producing truncated structures (Tn antigens and sialyl-Tn) that serve as disease biomarkers [ 17 ]. In ovarian, endometrial, and cervical cancers, aberrant glycosylation manifests through multiple converging mechanisms [ 10 ]. Enhanced branching of N-glycans through upregulation of GnT-V (encoded by MGAT5) creates structures that bind galectin-3, forming surface lattices that sequester growth factor receptors and drive constitutive proliferative signaling [ 18 ]. Hypersialylation—excessive terminal sialic acid residues—creates a “don’t eat me” glycan shield that engages Siglec receptors on immune cells, delivering inhibitory signals that suppress anti-tumor immunity [ 19 ]. Truncation of O-glycans, driven by epigenetic silencing of COSMC (a critical chaperone for O-glycan maturation), exposes tumor-associated carbohydrate antigens (TACAs) including sialyl-Tn (sTn), which are absent from healthy tissues but prevalent in gynecologic malignancies [ 20 ]. Polycystic ovary syndrome is characterized by insulin resistance, hyperandrogenism, and systemic inflammation, all of which are intimately linked to metabolic dysregulation and altered protein post-translational modifications [ 21 ]. The dysregulated glucose and lipid metabolism characteristic of PCOS results in altered availability of nucleotide sugar substrates (UDP-glucose, UDP-GlcNAc, CMP-sialic acid) used for glycosylation biosynthesis, thereby systematically reshaping the glycomic landscape of serum glycoproteins and reproductive tract tissues [ 22 ]. Specifically, enhanced glycolysis in PCOS shifts biosynthetic precursor pools toward glycolytic intermediates and away from oxidative pathways, creating a metabolic environment that favors particular glycan structures over others. Inflammatory cytokines elevated in PCOS (TNF-α, IL-6) activate inflammatory transcription factors that upregulate sialyltransferases and other glycan-modifying enzymes, resulting in elevated sialylation of circulating proteins [ 23 ]. These PCOS-associated glycomic signatures could serve as biomarkers for disease identification and stratification of disease severity, enabling earlier intervention and more precise therapeutic targeting [ 24 ]. Endometriosis is characterized by dysregulated estrogen signaling, elevated local inflammation, and altered endometrial receptivity, all processes intimately linked to protein modifications including glycosylation [ 25 ]. Estrogen receptor alpha (ESR1) and its coregulators, including ZMIZ1, control the expression of genes encoding glycosylation enzymes through transcriptional regulatory mechanisms [ 26 ]. In endometriosis, dysregulated estrogen signaling and reduced ZMIZ1 expression result in altered glycosylation patterns on proteins critical for cell-cell adhesion, immune cell recognition, and endometrial receptivity [ 27 ]. The identification of glycomic signatures specific to eutopic endometrium from women with endometriosis—distinguishing it from healthy controls—could facilitate non-invasive diagnosis and enable earlier intervention before ectopic lesion establishment [ 28 ].

The most clinically advanced glycomic biomarker application involves measuring glycan modifications on established cancer-associated proteins rather than developing entirely novel markers [ 29 ]. Recent 2024–2025 studies demonstrate that specific fucosylated and sialylated glycoforms of CA-125 achieve diagnostic accuracy for ovarian cancer exceeding 95% specificity, substantially surpassing the ~70% specificity of total CA-125 measurement [ 30 ]. This glycoform-specific approach represents a paradigm shift: rather than measuring crude protein abundance, clinicians assess the carbohydrate “language” encoded on these proteins, which differs characteristically between cancer and benign sources [ 31 ]. Cancer-derived CA-125 carries distinctive patterns of sialylation, fucosylation, and N-glycan branching that reflect the dysregulation of glycosyltransferases in tumor cells, creating recognizable molecular signatures absent from CA-125 produced by healthy endometrial or epithelial cells [ 32 ]. Traditional protein-only biomarkers, such as CA-125, often fail to distinguish between malignancy and benign inflammation, leading to high false-positive rates and unnecessary invasive procedures. The data presented here suggests that the solution lies in qualitative rather than quantitative analysis. By measuring the “carbohydrate language” of these proteins—specifically the patterns of hypersialylation and fucosylation—clinicians can achieve a level of specificity previously unattainable, distinguishing cancer-derived glycoproteins from their benign counterparts with high precision. Clinically, implementation of glycoform-specific biomarkers could function through reflex testing: when CA-125 is elevated using current thresholds, the sample is automatically analyzed for cancer-specific glycoforms [ 33 ]. If present, the probability of malignancy is high, warranting imaging and possible surgical intervention. If absent, the elevation likely reflects benign disease, avoiding unnecessary procedures. This simple addition to existing workflow infrastructure could dramatically reduce diagnostic uncertainty, unnecessary procedures, and patient anxiety while simultaneously improving cancer detection rates in early, more treatable stages [ 30 ]. Beyond free circulating glycoproteins, recent attention has shifted toward the glycomic profiling of extracellular vesicles (EVs), particularly exosomes. These membrane-bound nanovesicles secreted by tumor cells carry a unique molecular cargo that mirrors the parental cell’s status, including surface proteins with distinct tumor-associated glycosylation patterns [ 34 ]. In gynecological malignancies, specific alterations in N- and O-glycosylation on EV surfaces have emerged as critical diagnostic targets. For instance, ovarian cancer-derived exosomes have been shown to be enriched with specific sialoglycoproteins, such as Galectin-3-binding protein (LGALS3BP), and exhibit unique N-glycan signatures characterized by increased bisecting GlcNAc and α2,3-sialylation compared to benign vesicles [ 35 ]. These vesicular glyco-profiles offer a ‘concentrated’ source of tumor-specific markers, potentially overcoming the dilution and specificity issues inherent in analyzing whole serum [ 36 ]. The future of gynecological disease diagnostics lies in integrated panels combining glycomic, genomic, epigenomic, and metabolomic data within machine learning frameworks [ 37 ]. For gynecologic malignancies specifically, integrated biomarker panels might combine: (1) glycoform-specific CA-125 and HE4 measurements; (2) circulating tumor DNA (ctDNA) analysis revealing tumor-specific mutations and copy number alterations; (3) circulating tumor cell (CTC) enumeration and characterization; (4) extracellular vesicle-derived glycosylated exosomal microRNAs; and (5) circulating free DNA methylation patterns reflecting epigenetic changes [ 38 , 39 ]. Machine learning algorithms trained on prospective cohorts can integrate these multi-dimensional data to generate predictive models with a sensitivity and specificity far exceeding any single marker, achieving the diagnostic precision necessary for population-based screening or high-risk population surveillance. For PCOS diagnosis, integrated panels could combine classical biochemical parameters (testosterone, LH/FSH ratios) with genomic variants identified through genome-wide association studies (such as WNT4, ESR1, FSHB polymorphisms), with epigenetic biomarkers (DNA methylation patterns, miRNA expression profiles), and with novel glycomic signatures reflecting the metabolic and inflammatory state [ 40 ]. The OvAge algorithm, which integrates clinical, hormonal, and ultrasound parameters into a single output, exemplifies the successful integration of multiple data types to improve diagnostic accuracy and could be expanded to incorporate glycomic and other molecular data [ 41 ]. Uterine receptivity—the capacity of the endometrium to accept embryo implantation—is a precisely regulated physiological state controlled by estrogen and progesterone signaling through coordinated transcriptional and post-translational modifications of endometrial proteins [ 42 ]. Emerging evidence suggests that glycosylation of adhesion molecules, growth factors, and cytokines plays critical roles in regulating embryo–endometrial interactions essential for successful implantation [ 43 ]. Dysregulation of glycosylation patterns in the endometrium of women with recurrent implantation failure or recurrent pregnancy loss could serve as diagnostic biomarkers enabling the identification of women who would benefit from specific interventions to correct endometrial glycosylation defects [ 44 ]. The glycosylation of follistatin, a critical regulator of uterine receptivity through modulation of TGF-β family signaling, represents one example of glycomic involvement in reproductive competence [ 45 ]. Glycoform-specific quantification of follistatin and related molecules in serum, endometrial tissue, or follicular fluid could identify women at risk for implantation failure and could potentially guide targeted interventions to normalize glycosylation patterns [ 46 ]. Non-invasive diagnostic tests measuring endometrial glycomic signatures through liquid biopsy (circulating endometrial cells, extracellular vesicles, cfDNA methylation) could enable the preimplantation assessment of receptivity without invasive endometrial biopsy, transforming assisted reproductive technology and fertility preservation strategies [ 47 ].

Modern glycomic analysis relies upon high-resolution mass spectrometry techniques that enable the identification and quantification of individual glycan structures with extraordinary precision. HILIC-UPLC-MS/MS achieves the separation of glycan isomers—molecules with identical chemical formulas but different three-dimensional structures—a distinction critical since isomers frequently possess different biological functions and different associations with disease [ 48 ]. Matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI-MSI) enables the spatial mapping of glycan structures within tissue sections, revealing that specific glycan classes accumulate in particular tumor microenvironmental compartments such as cancer-associated fibroblasts, immune cell infiltrates, or hypoxic regions [ 49 ]. These spatial glycomic profiles provide insights into the glycomic landscape of the tumor microenvironment that could inform therapeutic targeting and predict immunotherapy responsiveness provide insights into the glycomic landscape of the tumor microenvironment that could inform therapeutic targeting and predict immunotherapy responsiveness [ 50 ]. The analytical standardization of glycomic assays remains an ongoing challenge requiring the development of standardized protocols, reference materials, proficiency testing programs, and harmonized data interpretation criteria [ 51 ]. Professional societies and consortia are establishing these standards, facilitating the translation of research-stage glycomic methods into clinically validated diagnostic tests. Quality control measures including instrument standardization, reagent lot standardization, and operator training will be essential for ensuring reproducibility across clinical laboratories [ 52 ]. The complexity of glycomic datasets—with hundreds to thousands of distinct glycan structures potentially present in a single sample—necessitates sophisticated computational approaches for pattern recognition and biological interpretation [ 53 ]. Machine learning algorithms including deep neural networks, random forests, support vector machines, and ensemble methods can identify multidimensional glycomic signatures predictive of disease state, disease severity, treatment response, or clinical outcomes [ 54 ]. Recent applications of convolutional neural networks (CNNs), particularly for imaging-based biomarker analysis using Raman spectroscopy combined with AI, have achieved diagnostic accuracies exceeding 95% for various cancer types, with similar applications emerging for gynecologic malignancies [ 55 ]. Critical to clinical adoption is the development of explainable AI approaches wherein the specific glycan features driving predictions are transparent to clinicians [ 56 ]. Black-box algorithms that achieve high accuracy but cannot be interpreted by clinicians will face substantial barriers to clinical implementation [ 57 ]. Rather, interpretable machine learning models that identify biologically meaningful glycomic patterns—for instance, elevated fucosylation specifically at core N-glycan positions combined with reduced sialylation on O-glycans—will enable clinical understanding and support integration into clinical decision-making workflows [ 58 ]. The recognition that cancer and gynecological diseases result from convergent dysregulation of multiple biological layers—genomic, epigenomic, transcriptomic, proteomic, and glycomic—has catalyzed the development of integrated multi-omics approaches that simultaneously measure mutations, epigenetic modifications, RNA expression, protein levels, and glycosylation patterns [ 59 ]. For instance, specific genomic mutations in gynecologic cancers (such as BRCA1/2 mutations, PTEN loss, or TP53 mutations) are associated with characteristic glycomic signatures reflecting altered metabolic state and signaling pathway activation downstream of these mutations [ 60 ]. Epigenetic modifications including DNA methylation patterns and histone post-translational modifications regulate the expression of glycosyltransferases and glycosidases, thereby controlling glycomic output [ 61 ]. Integration of these multi-layer datasets within machine learning frameworks enables the identification of convergent molecular signatures with substantially greater predictive power than any single data layer alone [ 62 ].

The convergence of diagnostic targets and therapeutic opportunities creates a unique “theranostic” landscape in women’s health. The same Sialyl-Tn (sTn) and Tn antigens that serve as highly specific biomarkers for ovarian and endometrial cancers are now being leveraged as targets for next-generation antibody–drug conjugates (ADCs) and CAR-NK therapies. Because these truncated O-glycans are essentially absent from healthy tissues, they offer a “narrow-spectrum” therapeutic window that could significantly reduce the off-target toxicities associated with conventional chemotherapy. However, despite this promise, significant hurdles remain. The complexity of the glycome requires a multidisciplinary approach, combining high-resolution analytical chemistry with clinical oncology to ensure that these glycan-targeted strategies can be standardized for routine clinical use. The same glycan structures that define disease through diagnostic biomarkers can be therapeutically targeted through immunotherapy approaches [ 63 ]. Antibody–drug conjugates (ADCs) directed against tumor-associated carbohydrate antigens (TACAs) such as sialyl-Tn (sTn) combine antibody-mediated specificity for tumor-restricted glycan epitopes with cytotoxic payloads, enabling the selective elimination of cancer cells while sparing healthy tissue. sTn-targeting ADCs are currently entering clinical trials for ovarian, endometrial, and cervical cancers, with preliminary data suggesting activity in disease resistant to conventional chemotherapy [ 64 ]. The theoretical advantage of sTn-targeting ADCs lies in the exquisite tumor selectivity: sTn is essentially absent on normal tissues yet abundant on gynecologic malignancies, potentially enabling therapeutic efficacy with substantially reduced off-target toxicity compared to conventional cytotoxic chemotherapy [ 65 ]. Chimeric antigen receptor (CAR)-engineered T cells and natural killer (NK) cells targeted against glycopeptide epitopes represent another rapidly advancing therapeutic platform [ 66 ]. CAR-NK cells targeting Tn-MUC1 (mucin-1 proteins carrying incomplete O-glycosylation with exposed Tn antigens) have demonstrated preliminary efficacy in early trials, with the advantage that engineered NK cells can be derived from allogeneic donors, potentially enabling “off-the-shelf” therapeutic products without requiring individualized manufacturing [ 67 ]. CAR-T cells and CAR-NK cells targeting glycopeptides combine the adoptive cellular immunity advantages of these platforms with the tumor selectivity conferred by glycan-specific targeting, offering potential for durable anti-tumor responses and possible long-term immune memory ( Table 2 ) [ 68 ]. Hypersialylation of tumor-derived proteins creates a “don’t eat me” glycan shield through engagement of Siglec receptors on immune cells, suppressing anti-tumor immunity through mechanisms functionally equivalent to protein-based checkpoint inhibition. Sugar-stripping enzymes, particularly sialidase fusion proteins that enzymatically remove terminal sialic acid residues, are currently entering clinical trials with the objective of unmasking hidden tumor antigens and eliminating glycan-mediated immune suppression [ 69 ]. The combination of sialidase-mediated glycan removal with conventional checkpoint inhibitors targeting protein-based checkpoints (such as anti-PD-1 antibodies) represents a rational therapeutic strategy targeting immune suppression through two complementary mechanisms [ 70 ]. Preclinical data support this concept, demonstrating synergistic anti-tumor activity when sialidase is combined with checkpoint inhibitors in multiple cancer models including gynecologic malignancies [ 71 ]. Direct inhibition of glycosyltransferases or glycosidases represents another therapeutic avenue for targeting glycomic dysregulation [ 72 ]. For instance, selective inhibition of sialyltransferases such as ST6GALNAC4, which is upregulated in many cancers through MYC-driven transcriptional activation, could reduce the hypersialylation that drives immune evasion [ 73 ]. Similarly, inhibitors of fucosyltransferases could reduce fucosylation-driven epithelial-to-mesenchymal transition and metastatic dissemination in cervical and other adenocarcinomas [ 74 ]. The therapeutic challenge lies in achieving sufficient selectivity to avoid disrupting the glycosylation of normal proteins in healthy tissues, potentially limiting clinical applicability [ 75 ]. However, the discovery that specific glycosyltransferases are preferentially upregulated in particular tumor types or are driven by specific oncogenic pathways (such as MYC-driven ST6GALNAC4 upregulation) offers opportunities for relative selectivity. Combination strategies pairing glycosyltransferase inhibition with conventional chemotherapy or immunotherapy could potentially improve outcomes by reducing glycan-mediated immune suppression while simultaneously treating the primary tumor [ 76 ].

Women’s gynecological health encompasses a complex spectrum of conditions ranging from metabolic and endocrine disorders to inflammatory diseases and malignancies [ 1 ]. Polycystic ovary syndrome (PCOS) affects 8–20% of reproductive-age women, disrupting metabolic homeostasis and reproductive function through dysregulated insulin signaling and hyperandrogenism [ 2 ]. Endometriosis, characterized by ectopic endometrial growth, affects 10–15% of reproductive-age women and is frequently misdiagnosed, with diagnostic delays averaging 7–10 years [ 3 ]. Gynecologic malignancies—ovarian, endometrial, and cervical cancers—collectively account for approximately 16.1% of new cancer diagnoses in women globally, with over 680,000 deaths annually, largely attributable to late-stage detection [ 4 ]. The fundamental challenge traversing all of these conditions is the inadequacy of current diagnostic biomarkers. For gynecologic malignancies, standard markers such as CA-125 (MUC16) and HE4 demonstrate sensitivity for early-stage disease of only 50–76% while generating substantial false positives in benign conditions [ 5 ]. For PCOS and endometriosis, diagnosis remains primarily clinical and imaging-based, lacking robust molecular biomarkers that could facilitate earlier intervention and personalized treatment selection [ 6 ]. These diagnostic limitations are compounded by the increasing recognition that traditional single-protein biomarkers capture insufficient biological information to distinguish disease states from healthy conditions or to predict individual therapeutic responsiveness. Glycosylation—the enzymatic attachment of carbohydrate chains (glycans) to proteins and lipids—represents the most abundant and chemically diverse post-translational modification in human cells, yet remains underutilized in clinical diagnostics [ 7 ]. More than 50% of human proteins undergo glycosylation, and remarkably, nearly all FDA-approved cancer biomarkers are glycoproteins [ 8 ]. The critical insight driving the glycomic revolution is that cancer cells and cells in pathological states exhibit systematic dysregulation of glycosylation enzymes (glycosyltransferases and glycosidases), resulting in characteristic carbohydrate modifications absent from healthy tissue [ 9 ]. These glycomic “signatures” represent a previously unexploited reservoir of diagnostic and therapeutic information that could transform precision medicine approaches across gynecological diseases [ 10 ]. Recent technological advances in mass spectrometry, artificial intelligence, and multi-omics integration have made glycomic analysis increasingly feasible for clinical implementation [ 11 ]. Techniques such as HILIC-UPLC-MS/MS, combined with machine learning algorithms and integrated with genomic, transcriptomic, and proteomic data, now enable the identification of glycomic signatures with unprecedented sensitivity and specificity [ 12 ]. This review synthesizes emerging evidence regarding glycomic dysregulation across gynecological diseases and demonstrates how glycomic biomarkers and glycan-targeting therapeutics represent a transformative approach to women’s health. Therefore, biomarkers and glycan-targeting therapeutics represent a transformative approach to women’s health. This review also identifies a distinct “metabolic–glycomic axis” that differentiates non-malignant conditions from cancers. While malignant glycomes are typically reshaped by epigenetic silencing of chaperones (e.g., COSMC ) or enzymatic reprogramming, the glycomic defects in PCOS and metabolic syndrome appear to be driven by substrate availability. The shift toward enhanced glycolysis in these patients alters the pool of nucleotide sugar donors, which systematically reshapes the serum glycome. This mechanistic distinction is critical for future research as it suggests that therapeutic interventions in metabolic gynecological disorders might focus on metabolic normalization to correct the glycome, whereas malignancies require the direct targeting of aberrant glycans or the immune checkpoints they engage. To facilitate a comprehensive understanding of this complex molecular landscape, Table 1 synthesizes the primary glycomic alterations observed across gynecological pathologies, distinguishing between the mechanisms driving malignancies and those underlying metabolic or inflammatory conditions.

Glycomics represents an emerging frontier in molecular diagnostics and therapeutics for gynecological diseases, offering unprecedented opportunities to improve women’s health through earlier detection, enhanced risk stratification, and personalized therapeutic targeting. The recognition that cancer cells and cells in pathological gynecologic states exhibit systematic dysregulation of glycosylation machinery, resulting in characteristic carbohydrate modifications absent from healthy tissue, has opened entirely new diagnostic and therapeutic avenues. Glycoform-specific biomarkers for established proteins achieve diagnostic specificity and sensitivity substantially exceeding protein concentration-based testing, promising to transform early cancer detection and reduce unnecessary procedures in women with benign disease. Integration of glycomic data with complementary omics technologies—genomic, epigenomic, transcriptomic, proteomic—and artificial intelligence-driven analysis creates a multi-dimensional molecular characterization of disease, enabling precision medicine approaches tailored to individual biology. Therapeutically, glycan-targeting strategies including antibody–drug conjugates, CAR-T cell therapy, and sugar-stripping enzymes represent genuinely novel mechanisms with activity in treatment-resistant disease. These approaches complement conventional therapies by targeting distinctive features of cancer and diseased cells—their aberrant glycosylation—offering potential for improved efficacy with potentially reduced off-target toxicity. Substantial work remains before widespread clinical implementation of glycomic diagnostics and therapeutics. Analytical standardization, large prospective clinical validation studies, regulatory approval, cost reduction, and the development of accessible technologies are all necessary next steps. However, the scientific foundation is solid, the clinical need is urgent, and the potential for transforming women’s health outcomes is substantial. Investment in glycomic research, multidisciplinary collaboration, and translational research initiatives linking molecular discoveries to clinical applications will accelerate the integration of glycomics into precision medicine for gynecological diseases. As these advances progress over the next 2–3 years, gynecologic clinicians should anticipate that glycomic biomarkers will increasingly influence early detection algorithms, risk stratification, treatment selection, and treatment monitoring, fundamentally improving outcomes for women with gynecologic disorders globally.