{"paper_id":"4bbf8f2d-5d0f-4588-b139-312f27e32ca4","body_text":"Vol.:(0123456789)\nJournal of Applied Genetics \nhttps://doi.org/10.1007/s13353-025-01021-y\nHUMAN GENETICS • REVIEW\nPotential biomarkers for early detection of endometriosis: current \nstate of art (what we know so far)\nMichalina Kliber‑Gałuszka1 · Klaudia Kulczyńska‑Figurny1  · Paweł Piotr Jagodziński1  · Andrzej Pławski2\nReceived: 20 June 2025 / Revised: 2 September 2025 / Accepted: 23 September 2025 \n© The Author(s) 2025\nAbstract\nEndometriosis is a chronic gynecological condition characterized by the presence of endometrial-like tissue outside the \nuterine cavity. Its diagnosis remains a significant clinical challenge, often delayed by 7 to 12 years, leading to considerable \nsocio-economic burden and a substantial decline in patients’ quality of life, including potential infertility. Consequently, \nthere is an urgent need to identify reliable biomarkers that would allow for earlier and more accurate detection. This review \nprovides a comprehensive and up-to-date analysis of potential biomarkers for the diagnosis of endometriosis, including \nhormonal, inflammatory, genetic, epigenetic, immunological, metabolic, and imaging-based markers. Their diagnostic value \nand limitations are critically assessed, with particular emphasis on the advantages of multimarker and integrated diagnostic \napproaches to enhance early detection. The findings of this review offer valuable insights for clinicians, researchers, and \nhealthcare professionals working to develop better diagnostic methods and improve patient outcomes. Moreover, the integra-\ntion of emerging technologies, such as artificial intelligence, offers promising opportunities to revolutionize endometriosis \ndiagnostics through personalized and precise medical care.\nKeywords Endometriosis · Biomarkers · Artificial intelligence · Machine learning · Personalized medicine\nAbbreviations\nAGD  Anogenital distance\nAI  Artificial intelligence\nAUC   Area under the curve\nHE4  Human epididymis protein 4\nMRI  Magnetic resonance imaging\nML  Machine learning\nTGFBI  Transforming growth factor Beta-induced \nprotein\nTNF alpha  Tumor Necrosis Factor-alpha\nTVUS  Transvaginal ultrasound\nIntroduction\nEndometriosis is an estrogen-dependent inflammatory \ndisease defined by the presence of endometrial-like tissue \noutside the uterine cavity, affecting approximately 10% of \nwomen of reproductive age worldwide (Bulun et al. 2019). \nAlthough endometriosis affects a significant portion of the \nfemale population, the path to diagnosis is often prolonged, \nwith many patients enduring years of misdiagnoses and \nclinical uncertainty before receiving a correct identifica -\ntion of the disease—typically delayed by 7 to 12 years from \nsymptom onset (Swift et al. 2024).\nThe diagnostic challenges are compounded by the hetero-\ngeneity of its clinical manifestations—ranging from severe \npelvic pain, dysmenorrhea, and infertility to entirely asymp-\ntomatic presentations—often mimicking other gynecologi-\ncal or gastrointestinal disorders (Murphy 2002; Lukac et al. \n2022).\nThis enigmatic disorder not only has profound implica-\ntions for women’s physical and emotional well-being but \nalso imposes a substantial socio-economic burden. Women \nwith endometriosis often experience reduced quality of life, \ndue to chronic pain, infertility, and the psychological toll of \nliving with a long-term condition. These challenges extend \nCommunicated by: Kamila Kusz-Zamelczyk\n * Klaudia Kulczyńska-Figurny \n kkulczynska-figurny@ump.edu.pl\n1 Department of Biochemistry and Molecular Biology, \nPoznan University of Medical Sciences, Święcickiego 6 St., \n61-701 Poznan, Poland\n2 Institute of Human Genetics, Polish Academy of Sciences, \nStrzeszyńska 32 St., 60-479 Poznan, Poland\n\n Journal of Applied Genetics\nbeyond the individual to the broader society, contributing to \nsubstantial healthcare costs and work absenteeism. A study \nby Simoens et al. (2012) estimated that the annual total cost \nof treating a woman with endometriosis in referral centres \nwas €9579, with €6298 attributed to lost productivity and \n€3113 to healthcare costs. The burden was found to increase \nwith the severity of the disease, the presence of pelvic pain, \ninfertility, and the duration of the diagnostic delay, further \nemphasizing the significant socio-economic implications of \nthis condition (Simoens et al. 2012).\nCurrently, the gold standard for diagnosis remains lapa-\nroscopic surgery with histological confirmation, an inva-\nsive approach that underscores the pressing need for non-\ninvasive, reliable diagnostic alternatives (Simko and Wright \n2022).\nEndometriosis presents a unique conundrum in the realm \nof biomarkers—it is a condition that manifests in diverse \nways, affecting women of varying ages, and may be accom-\npanied by a spectrum of symptoms, from the debilitating to \nthe entirely asymptomatic.\nIn recent years, advances in molecular biology and clini-\ncal diagnostics have spurred growing interest in the identifi-\ncation of disease-specific biomarkers. Biomarkers—defined \nas measurable indicators of normal or pathological biologi-\ncal processes—hold the potential to transform the diagnos-\ntic landscape of endometriosis (Encalada Soto et al. 2022). \nThey could facilitate earlier detection, predict disease pro-\ngression, assess therapeutic responses, and ultimately pave \nthe way for personalized treatment strategies.\nAmong the most promising areas of biomarker discovery \nin endometriosis are genetic and epigenetic factors. Family \nstudies and twin concordance rates indicate a strong heredi-\ntary component, with an estimated heritability of up to 50%. \nGenome-wide association studies (GWAS) have identified \nmultiple risk loci—including SNPs in genes such as WNT4, \nVEZT, and GREB1—that are consistently associated with \nendometriosis across populations (Nyholt et al. 2012). Addi-\ntionally, epigenetic modifications such as DNA methylation \npatterns and dysregulated miRNA expression have emerged \nas crucial contributors to disease onset, lesion development, \nand hormonal resistance. These molecular alterations not \nonly enhance our understanding of disease pathophysiol -\nogy but also offer a promising foundation for non-invasive \ndiagnostic strategies and polygenic risk modeling.\nRecent advances in biomarker research have shown prom-\nising potential in revolutionizing the diagnosis of this com-\nplex disorder. Biomarkers, defined as objective and quan-\ntifiable indicators of biological processes, offer hope for \nthe early detection of endometriosis, assessment of disease \nseverity, and monitoring of its progression.\nThis review provides an in-depth evaluation of cur -\nrent and emerging biomarkers in endometriosis research, \nincluding hormonal, inflammatory, genetic, epigenetic, \nimmunological, metabolic, and imaging-related mark -\ners. Each category reflects a distinct biological dimension \nof the disease and contributes to a more nuanced under -\nstanding of its complex pathophysiology. In addition, the \npotential of multimarker panels and integrated diagnostic \nplatforms—including artificial intelligence and omics-based \napproaches—will be explored, offering a vision of future \nprecision medicine applications in endometriosis care.\nMoreover, the integration of artificial intelligence (AI) \nand machine learning (ML) into the diagnostic process holds \nenormous potential (Hanna et al. 2025). AI algorithms, \ntrained on vast datasets of patient information, could assist \nin identifying patterns and correlations that may not be \napparent to the human eye. ML models, in particular, are \nwell-suited for analyzing complex, multidimensional data \nsuch as biomarker profiles, imaging results, and clinical his-\ntories. These models could help clinicians predict disease \nprogression, assess treatment responses, and even identify \npersonalized therapeutic strategies tailored to individual \npatients (Bekbolatova et al. 2024).\nThe combination of AI and biomarkers has the poten-\ntial to revolutionize not only the detection of endometrio-\nsis but also its management. By integrating omics-based \napproaches—such as genomics, proteomics, and metabo-\nlomics—into AI-driven diagnostic platforms, a more com-\nprehensive understanding of the disease could emerge. \nThese platforms could enable clinicians to offer person-\nalized, precision medicine approaches, moving beyond a \none-size-fits-all model to a more tailored treatment strategy \nthat optimally addresses the unique characteristics of each \npatient's condition. As we move forward, the continuous \nevolution of AI, ML, and omics technologies promises to \nreshape the landscape of endometriosis diagnosis and care, \noffering women earlier diagnoses, more effective treatments, \nand a better overall quality of life (Dungate et al. 2024).\nTypes of biomarkers\nTo date, researchers have investigated various types of bio-\nmarkers for the diagnosis of endometriosis, including hor -\nmonal, inflammatory, genetic, epigenetic, immunological, \nmetabolic, and imaging biomarkers (Fig. 1 ). However, cur-\nrently no single biomarker can diagnose endometriosis with \nhigh accuracy and specificity. The combination of these bio-\nmarkers, especially when supported by advanced technolo-\ngies such as machine learning, offers new opportunities for \nimproved early detection.\nHormonal biomarkers\nHormonal biomarkers have long been central to under -\nstanding endometriosis, as this condition exhibits hormonal \n\nJournal of Applied Genetics \ndependencies. In addition to classical hormones such as \nestrogen, progesterone, LH, and FSH, increasing attention \nhas been paid to enzymes and metabolites involved in estro-\ngen metabolism, which may serve as diagnostic and prog-\nnostic biomarkers.\nElevated estrogen levels and disrupted hormone ratios \nhave been identified as potential indicators of endometriosis. \nAromatase (CYP19A1), the enzyme responsible for convert-\ning androgens into estrogens, shows increased expression in \nendometrial tissues of patients with endometriosis. A meta-\nanalysis including 17 studies and 1279 participants demon-\nstrated that aromatase had the highest diagnostic accuracy \namong evaluated hormonal biomarkers, with a pooled sensi-\ntivity of 79% and specificity of 89%, outperforming estrogen \nreceptors (ERα/β), serum prolactin, and 17β-hydroxysteroid \ndehydrogenase type 2 (17βHSD2) (Gao et al. 2019).\nRecent findings indicate that the expression levels of \naromatase, steroidogenic factor-1 (SF-1), and HSD17B2 \nin menstrual blood could effectively differentiate patients \nwith endometriosis from healthy controls, with aromatase \nachieving an area under the curve (AUC) of 0.977 (Amanda \net al. 2024).\nElevated levels of 2-hydroxyestradiol (2OHE2) and \n4-hydroxyestradiol (4OHE2) have been reported in the \neutopic endometrium of women with endometriosis com-\npared to controls (Othman et al. 2021). Moreover, urinary \nconcentrations of 2-hydroxyestrone (2OHE1) are signifi-\ncantly higher in affected individuals, suggesting potential for \nnon-invasive diagnostic applications (Othman et al. 2021).\nAdditionally, recent studies have identified overexpres-\nsion of nicotinamide N-methyltransferase (NNMT) in endo-\nmetrial stromal cells, induced by estrogen and macrophage \ninteraction, which modulates cell proliferation via the \nNNMT-ERBB4-PI3K/AKT signaling pathway, contribut-\ning to the development of endometriosis (Hou et al. 2024).\nProgesterone resistance is another key hormonal aspect of \nendometriosis. Reduced expression of progesterone recep-\ntors and disrupted signaling pathways have been impli-\ncated in lesion persistence. Decreased FKBP4 levels and \nalterations in microRNA-29c regulation have been linked \nto impaired progesterone responsiveness (Joshi et al. 2017) . \nMoreover, studies have shown that loss of progesterone \nreceptor-B (PR-B) in stromal cells of endometriotic lesions \nis a hallmark of progesterone resistance, contributing to \ninfertility in endometriosis patients. Dual inhibition of AKT \nand ERK1/2 pathways has been proposed to restore pro-\ngesterone responsiveness in these cases (Dutta et al. 2018).\nRecent studies suggest that circulating testosterone, par -\nticularly in its free and bioavailable forms, may serve as \na potential biomarker for endometriosis diagnosis and risk \nstratification (Zhao et al. 2023; McGrath et al. 2023; Gjor-\ngoska and Rizner 2024). Lower levels of total testosterone, \nFig. 1  Proposed categories of \npotential biomarkers for early \ndetection of endometriosis, \nincluding hormonal, inflamma-\ntory, immunological, metabolic, \ngenetic/epigenetic, and imaging \nmarkers, which together illus-\ntrate the multifactorial nature of \nthe disease. Part of the illustra-\ntion was adapted from Servier \nMedical Art, licensed under a \nCreative Commons Attribu-\ntion 4.0 International License \n(https:// smart. servi er. com)\n\n\n Journal of Applied Genetics\nbioavailable testosterone, and DHEAS have been geneti-\ncally linked to a higher risk of endometriosis, highlighting \na potential hormonal profile characteristic of the disease \n(McGrath et al. 2023).\nFurthermore, reduced testosterone concentrations in fol-\nlicular fluid observed in endometriosis patients undergoing \nassisted reproductive technologies suggest its utility as a \nnon-invasive biomarker for impaired folliculogenesis and \ninfertility in these individuals (Huang et al. 2021). Although \ncurrent findings are inconsistent, integrating testosterone \nlevels with other hormonal, genetic, and inflammatory \nmarkers could enhance the sensitivity and specificity of \nbiomarker panels for early detection of endometriosis.\nInflammatory biomarkers\nEndometriosis is characterized by chronic inflammation, \nwith recent studies highlighting its role in the pathophysiol-\nogy of the disease. This persistent inflammatory response \ncontributes to pain, infertility, and lesion progression, and \nserves as the foundation for exploring inflammatory bio-\nmarkers in early diagnosis. Biomarkers in endometriosis \nplay a significant role in understanding its pathophysiology \nand in the development of non-invasive diagnostic tests.\nCytokines, both pro-inflammatory and anti-inflammatory, \nhave been identified in biopsy specimens from individuals \nwith endometriosis, suggesting their involvement in the dis-\nease's etiopathogenesis, similarly to what has been observed \nin various cancers (AlAshqar et al. 2021). For instance, mac-\nrophage migration inhibitory factor (MIF), alongside inter -\nleukin-1 (IL-1), has been found to play a regulatory role in \nimmune responses, angiogenesis, and estrogen production, \nall of which are critical in the progression of endometriosis \n(Cao et al. 2005; Veillat et al. 2010).\nFurthermore, cytokines such as IL-6, IL-8, and Tumor \nNecrosis Factor-alpha (TNF-alpha) are implicated in the \npain, embryonic implantation, and angiogenesis associ-\nated with endometriosis. However, these cytokines still \nrequire validation regarding their specificity, sensitivity, \nand diagnostic significance (Weisheng et al. 2019). Proin-\nflammatory cytokines, such as IL-1β (Taketani et al. 1992; \nHo et al. 1996), TNF-α (Overton et al. 1996; Harada et al. \n1997), IL-6 (Punnonen et al. 1996; Harada et al. 1997), and \nIL-8 (Arici et al. 1996) are elevated in the peritoneal fluid \nof women with endometriosis. They are secreted by perito-\nneal macrophages and ectopic endometrial lesions. These \ncytokines promote the development of endometriosis by \ninducing COX-2 expression and triggering PGE2 produc-\ntion, creating a positive feedback loop (Wu et al. 2002; Carli \net al. 2009). In particular, IL-1β upregulates COX-2 mRNA \nstability and promoter activity in ectopic endometrial cells, \nenhancing their migration and invasiveness (Taketani et al. \n1992; Ho et al. 1996). Similarly, TNF-α stimulates IL-6 and \nIL-8 secretion, further driving inflammation and the forma-\ntion of endometriotic lesions (Overton et al. 1996; Harada \net al. 1997).\nNon-invasive diagnostic approaches utilizing antibody \narrays have shown that IL-31 could be a potential biomarker, \nthough traditional markers like CA-125 lack the necessary \nsensitivity and specificity for accurate diagnosis (Waelkens \net al. 2020).\nAdditionally, pro-inflammatory cytokines like IL-17 and \nIL-33, commonly associated with both endometriosis and \ncardiovascular diseases, further complicate the development \nof reliable biomarkers (Rafi et al. 2021). Oxidative stress is \nanother important factor in the progression of endometriosis, \nwith damage-associated molecular patterns (DAMPs) like \nHMGB1 and TLR4 being identified as key components in \nthe inflammatory response associated with the disease (Yun \net al. 2016).\nMoreover, cytokines such as IL-6 and IL-10, along with \nfactors like TNF-alpha, are involved in the growth of endo-\nmetriotic cells and the potential carcinogenic effects of \nendometriosis (Wang et al. 2018) . However, as studies con-\ntinue, more evidence is needed to confirm the full potential \nof these cytokines as reliable biomarkers for diagnosis and \ntherapeutic targeting in endometriosis.\nAs inflammation is closely intertwined with immune dys-\nregulation, further exploration of immunological markers \nmay offer additional insights into the underlying mecha-\nnisms and diagnostic potential in endometriosis.\nImmunological biomarkers\nThe immune system plays a crucial role in the pathogen-\nesis of endometriosis. It has been proposed that defects in \nimmune surveillance and clearance mechanisms permit \nendometrial cells to survive, implant, and proliferate outside \nthe uterine cavity. The frequent coexistence of endometriosis \nwith autoimmune diseases, such as systemic lupus erythe-\nmatosus or Hashimoto’s thyroiditis, supports the hypoth-\nesis of underlying immune dysregulation in this condition \n(Blanco et al. 2025).\nAlterations in both innate and adaptive immune responses \nhave been documented in women with endometriosis. \nAmong innate immune cells, macrophages exhibit a skewed \npolarization toward the M2 phenotype, contributing to tissue \nremodeling, angiogenesis, and immune tolerance. Addition-\nally, natural killer (NK) cells display reduced cytotoxicity \nand altered cytokine secretion profiles, impairing their abil-\nity to eliminate ectopic endometrial cells (Jeung et al. 2016). \nDysregulated populations of dendritic cells and neutrophils \nhave also been observed, further implicating innate immu-\nnity in the persistence of endometriotic lesions (Jeung et al. \n2016).\n\nJournal of Applied Genetics \nFrom the adaptive immune perspective, changes in T lym-\nphocyte subsets, including an imbalance between Th1/Th2 \nand Th17/Treg cells, have been reported in both peritoneal \nfluid and peripheral blood (Tarokh et al. 2019; Pashizeh \net al. 2020; Olkowska-Truchanowicz et al. 2021). These \nalterations may favor a pro-inflammatory environment con-\nducive to lesion survival and progression. B cells, though \nless extensively studied, are also involved and may contrib-\nute by producing autoantibodies associated with endome-\ntriosis (Riccio et al. 2017; Harden et al. 2023).\nSeveral studies have identified circulating autoantibod-\nies targeting specific autoantigens, such as stomatin-like \nprotein 2 (SLP2), tropomodulin 3 (TMOD3), tropomyosin \n3 (TPM3), and PDIK1L. These autoantibodies have been \nparticularly associated with early-stage endometriosis and \nmay serve as promising non-invasive biomarkers for early \ndetection (Gajbhiye et al. 2017).\nAnother component of the immune system implicated in \nendometriosis is the complement system. Aberrant activa-\ntion and increased expression of complement factors such as \nC3 and C5 have been observed in the peritoneal fluid and tis-\nsues of affected individuals. These molecules may contribute \nto chronic inflammation, angiogenesis, and immune evasion \nby ectopic endometrial cells (Rahal et al. 2021).\nDysregulation of cell adhesion molecules, such as integ-\nrin β3 (CD61), has also been observed in the endometrium \nof women with endometriosis. Abnormal expression of these \nmolecules may facilitate ectopic implantation and lesion per-\nsistence (May et al. 2011).\nFrom a clinical perspective, immunological biomarkers \noffer potential utility not only for diagnosis but also for dis-\nease staging and therapeutic targeting. Recent transcriptomic \nand proteomic analyses have highlighted several immune-\nrelated genes and pathways—including CXCL12, PECAM1, \nNGF, CTGF, and WNT5A—that are differentially expressed \nin endometriotic lesions and may serve as future therapeu-\ntic targets (Yang et al. 2023; Zhang et al. 2025). Although \nimmunological markers are still under investigation and \nrequire validation in large, diverse cohorts, their integra-\ntion into multimodal diagnostic strategies could improve the \nearly detection, stratification, and personalized management \nof endometriosis.\nGenetic and epigenetic biomarkers\nRecent advances in genetic and epigenetic research have \ngreatly advanced the understanding of endometriosis, \nrevealing key molecular markers and mechanisms that con-\ntribute to the disease's pathogenesis. Studies have identi -\nfied several critical genes associated with endometriosis, \nincluding CUX2, CLMP, CEP131, EHD4, CDH24, ILRUN, \nLINC01709, HOTAIR, SLC30A2, and NKG7, which have \nbeen proposed as potential biomarkers for diagnosis and \ntreatment. These findings were supported by machine \nlearning-based approaches using transcriptomic datasets \nfrom patients with endometriosis and healthy controls. The \nBagged CART model, for example, achieved high classifi-\ncation metrics, including 85.7% accuracy, 100% sensitivity, \nand 75% specificity (Kucukakcali et al., 2025).\nIn addition to genetic factors, epigenetic modifications \nsuch as DNA methylation have been shown to play a crucial \nrole in the disease. A recent study identified 51 methylation \nquantitative trait loci (mQTLs) associated with endometrio-\nsis susceptibility, distributed across 21 genomic loci. These \nmQTLs offer valuable insights into tissue-specific epigenetic \nregulation and its potential impact on the development of the \ndisease (Mortlock et al. 2023). Another recent systematic \nreview further emphasized the role of DNA methylation pat-\nterns in the pathogenesis of endometriosis, suggesting that \nthe identification of specific methylation markers could sup-\nport novel diagnostic and therapeutic approaches (Ducreux \net al. 2025).\nMicroRNAs (miRNAs) have also been found dysregu-\nlated in endometriosis, potentially contributing to proges-\nterone resistance. MiR-199a-3p, miR-1-3p, miR-146a-5p, \nand miR-125b-5p were upregulated in ectopic lesions com-\npared to eutopic tissue (Hon et al. 2024). At the same time, \nERα and ERβ expression was altered, while PR-A and PR-B \nlevels remained unchanged. Predicted target genes of these \nmiRNAs (e.g., SCD, CDK6, DDIT4) are involved in pro-\nliferation and survival pathways, suggesting that miRNA-\ndriven hormonal dysregulation plays a key role in endome-\ntriosis pathogenesis and resistance to progestins (Hon et al. \n2024).\nMoreover, several miRNAs have been consistently identi-\nfied as promising biomarkers for non-invasive diagnostics. \nThese include miR-17-5p, miR-451a, and let-7b-5p, which \nhave consistently been found to be dysregulated across mul-\ntiple studies (Vanhie et al. 2024). New studies also propose \nthat differences in miRNA expression can be detected in \nbody fluids, offering opportunities for non-invasive diag-\nnostic tests (Ravaggi et al. 2024), and that salivary miRNA \nsignatures might help identify peritoneal superficial endo-\nmetriosis (Bendifallah et al. 2024).\nEndometriosis is a complex, estrogen-dependent condi-\ntion characterized by the ectopic growth of endometrial-\nlike tissue. It exhibits a strong heritable component, with \nheritability estimated at approximately 50%. Large-scale \ngenome-wide association studies (GWAS) have identified \nmultiple single nucleotide polymorphisms (SNPs) associated \nwith increased disease risk. These findings open the door to \nearly risk stratification and precision diagnostics.\nA landmark meta-analysis by Sapkota et al. (2017) uncov-\nered several key SNPs. Among them, rs12700667 on chro-\nmosome 7p15.2 showed strong and consistent association \nwith all disease stages. Though intergenic, this variant likely \n\n Journal of Applied Genetics\nregulates nearby gene expression and has been replicated \nacross diverse populations. Another key SNP, rs2235529, \nresides within the WNT4 gene (1p36.12), a major regulator \nof female reproductive tract development and decidualiza-\ntion. Dysregulation of WNT4 may impair tissue differentia-\ntion, contributing to lesion formation (Sapkota et al. 2017).\nAdditional risk variants include rs10859871 near the \nVEZT gene (12q22), which influences cell adhesion and \nepithelial barrier function—both relevant to the invasive \nnature of endometriotic lesions. Furthermore, rs13394619 \n(Nyholt et al. 2012) near GREB1 (2p25.1), a gene critical \nfor estrogen receptor signaling, has been implicated in dis-\nease susceptibility. Finally, variants near FN1 (fibronectin \n1) at 2q35, a gene involved in extracellular matrix remod-\neling, may drive tissue invasion and fibrosis characteristic \nof advanced endometriosis (Lalami et al. 2021).\nIn summary, the most promising SNPs identified across \nrecent studies include:\n• rs12700667 (7p15.2) – intergenic, consistently associated \nwith endometriosis risk.\n• rs2235529 (WNT4, 1p36.12) – involved in reproductive \ntract development.\n• rs10859871 (VEZT, 12q22) – affects cell adhesion and \nepithelial structure.\n• rs13394619 (GREB1, 2p25.1) – modulates estrogen \nresponse.\n• Variants near FN1 (2q35) – associated with tissue remod-\neling and lesion invasiveness.\nThese discoveries not only provide insights into the \nmolecular basis of endometriosis but also offer promising \ncandidates for integration into polygenic risk scores. With \nfurther functional validation, these SNPs could serve as the \nfoundation for non-invasive genetic screening tools, ena-\nbling earlier diagnosis and personalized management of \nendometriosis (Table  1).\nIn addition to GWAS-identified SNPs, recent transcrip-\ntomic and machine learning approaches have uncovered \nadditional candidate genes that may serve as promising \nnon-invasive biomarkers. These include CUX2, a tran-\nscription factor involved in cell cycle regulation and \ntissue remodeling, which has been shown to be upregu-\nlated in endometriotic lesions (Fan et al. 2021). CLMP, \na cell adhesion molecule, has been implicated in disrup-\ntion of cell–cell junctions and immune-related pathways, \npotentially facilitating ectopic implantation of endometrial \ntissue. CEP131, a centrosomal protein essential for cili-\nogenesis and cell cycle control, was also identified as dif-\nferentially expressed, pointing toward dysregulated signal-\ning and division in disease progression. EHD4, associated \nwith endocytic recycling and vesicular transport, may con-\ntribute to altered membrane dynamics and lesion survival.\nAmong genes implicated in adhesion and invasion, \nCDH4 (cadherin-4) regulates epithelial–mesenchymal \ntransition (EMT), a process critical for lesion estab-\nlishment, and its dysregulation has been proposed as \na biomarker of aggressiveness (Sapkota et al. 2017). \nImmune-related genes such as ILRUN, a regulator of type \nI interferon responses, and NKG7, involved in NK cell \ndegranulation, highlight immune dysregulation as a central \nmechanism; downregulation of NKG7 in particular corre-\nlates with impaired NK cell cytotoxicity (Fan et al. 2021).\nNon-coding RNAs also show strong biomarker poten-\ntial. LINC01709, a long non-coding RNA, has been sug-\ngested as a diagnostic marker due to its dysregulation in \nendometriotic samples, while HOTAIR, a well-character -\nized oncogenic lncRNA, promotes invasion, proliferation, \nand epigenetic reprogramming in ectopic endometrium. \nGiven their stability in circulation, both may serve as \nattractive candidates for minimally invasive assays (Sap-\nkota et al. 2017 ; Fan et al. 2021). Finally, SLC30A2N, \na zinc transporter-related gene, reflects perturbations in \nmetal ion homeostasis, which may contribute to oxidative \nstress and inflammatory signaling in lesions.\nCollectively, these candidate genes—CUX2, CLMP, \nCEP131, EHD4, CDH4, ILRUN, LINC01709, HOTAIR, \nSLC30A2N, and NKG7—represent a promising panel \nof molecular biomarkers identified through integrative \ngenomic and transcriptomic analyses. Their validation in \nlarge, multicenter cohorts could provide the basis for accu-\nrate, non-invasive diagnostic tools for endometriosis (Fan \net al. 2021; Zondervan et al. 2020; Sapkota et al. 2017.\nTable 1  The table summarizes selected SNPs linked to endometriosis risk, including their chromosomal loci, genomic regions, and odds ratios \nwith 95% confidence intervals\nSNP Locus (ChR) Genomic region/gene Odds (95% CI) Source (Citation)\nrs12700667 7p15.2 Intergenic (probable regulatory) 1.20 (1.13–1.27) (Nyholt et al. 2012)\nrs2235529 1p36.12 WNT4 (LINC00339–WNT4 locus) 1.29 (1.18–1.40) (Albertsen et al. 2013)\nrs10859871 12q22 Near VEZT (Not reported in source) (Nyholt et al. 2012)\nrs13394619 2p25.1 GREB1 (intronic) 0.92 (0.88–0.96) (Sapkota et al. 2017)\nrs1250248 (proxy FN1) 2q35 FN1 1.87 (1.34–2.61) (Matalliotaki et al. 2019)\n\nJournal of Applied Genetics \nBoth intergenic and gene-associated variants are shown, \nreflecting their potential roles in gene regulation, reproduc-\ntive tract development, extracellular matrix remodeling, and \nestrogen signaling\nMetabolic and mitochondrial biomarkers\nAlterations in metabolic pathways have been observed in \nwomen with endometriosis. Metabolomic studies—utilizing \nblood, urine, and tissue samples—have identified changes \nin the levels of specific metabolites, offering insight into \ndisease-related metabolic disturbances and highlighting their \npotential diagnostic and therapeutic relevance.\nMetabolomic analyses have revealed specific alterations \nin biochemical pathways among affected individuals. For \ninstance, elevated serum levels of amino acids such as leu-\ncine, lysine, alanine, valine, tyrosine, and phenylalanine \nhave been associated with the presence of endometriosis, \nas reported by Wang et al. (2018) (Dutta et al. 2018). Addi-\ntionally, disruptions in purine metabolism have also been \nnotedincreased levels of hypoxanthine inosine guanosine \nand xanthosine decreased concentrations of uric acid. These \nchanges suggest altered nucleotide Turnover and enhanced \ncellular stress, as described by Li et al 2018 (Li et al. 2018).\nThe diagnostic utility of these metabolic markers has \nshown promise. For example, a diagnostic model combin-\ning hypoxanthine, uric acid, and lysophosphatidylethanola-\nmine achieved a sensitivity of 66.7% and specificity of 90% \nin detecting early-stage endometriosis. Another biomarker \npanel, including 3-hydroxybutyrate, threonic acid, and ala-\nnine, yielded an area under the curve (AUC) of 0.91, indicat-\ning high diagnostic accuracy (Li et al. 2018).\nLipidomic profiling has further demonstrated that patients \nwith endometriosis show increased levels of lysophosphati-\ndylethanolamine and omega-3 arachidonic acid in endome-\ntrial tissues, indicating dysregulation of membrane Lipid \ncomposition and inflammatory Lipid mediators. This was \nexplored in detail by Ortiz et al 2021 (Ortiz et al. 2021). \nAlthough these findings are encouraging, further validation \nthrough large-scale, multicenter studies is essential. Con -\nfirming the clinical applicability of metabolic biomarkers \ncould pave the way for their integration into routine diag-\nnostics, particularly when combined with other omics-based \napproaches to enhance diagnostic sensitivity and specificity.\nImaging and anthropometric biomarkers\nImaging biomarkers, including anthropometric indicators \nand advanced radiological techniques, have gained (Zapar -\ndiel et al. 2016)increasing attention in the context of endo-\nmetriosis diagnostics, providing non-invasive alternatives \nor complementary tools to laparoscopy. Anthropometric \nmarkers such as the second-to-fourth digit ratio known as \n(2D:4D) have been studied as indicators of prenatal hor -\nmonal exposure. Numerous studies have demonstrated that \nwomen with endometriosis tend to have a higher 2D:4D \nratio, particularly on the right hand, which may reflect \nlower prenatal androgen exposure and higher estrogen lev -\nels—both of which are associated with increased disease \nsusceptibility, as shown by Ribeiro et al. (2023) (Buggio \net al. 2023).\nAnother antropometric biomarker—anogenital distance \n(AGD)—is a significant marker that reflects prenatal andro-\ngen exposure and has been studied as a potential diagnostic \nbiomarker for endometriosis. AGD can be measured using \ndifferent urogenital landmarks. For instance,  AGDAC refers \nto the distance from the anterior surface of the clitoris to the \nupper edge or center of the anus, while  AGDAF is measured \nfrom the posterior fourchette to the upper edge or center of \nthe anus, and  AGDCt is taken from the tip of the clitoris to \nthe center of the anus (Mendiola et al. 2012).\nRecent studies, including a meta-analysis and systematic \nreview, have suggested that a shorter  AGDAF may be a prom-\nising biomarker for endometriosis diagnosis. However, the \nresults remain inconsistent, and further research is required. \nFor example, Crestani et al. (2023) found that both  AGDAC \nand  AGDAF were significantly shorter in women with endo-\nmetriosis in a French population (Crestani et al. 2020). In \ncontrast, Peters et al. (2021) in the Netherlands observed \nthat only  AGDAC was shorter in women with endometrio-\nsis compared to controls (Peters et al. 2020). Conversely, a \nstudy in Spain showed that only  AGDAF was associated with \nendometriosis (Sánchez-Ferrer et al. 2017). These discrepan-\ncies highlight the need for more extensive studies to better \nunderstand the role of AGD in diagnosing endometriosis.\nAdditionally, birth weight has been explored as a poten-\ntial biomarker, with population studies suggesting that \nwomen with low birth weight are at increased risk of devel-\noping endometriosis later in life. For example, Borghese \net al. (2015), in a study involving 743 women, found that \nthose with histologically confirmed endometriosis had sig-\nnificantly lower average birth weights. Moreover, a birth \nweight below 2500 g was associated with a higher risk of \ndeep infiltrating endometriosis, suggesting that intrauterine \ngrowth restriction may contribute to the disease’s etiology \n(Borghese et al. 2015).\nBeyond these developmental markers, imaging techniques \nsuch as transvaginal ultrasound (TVUS) remain the first-\nline diagnostic tool in suspected endometriosis cases, par -\nticularly for identifying ovarian endometriomas and deeply \ninfiltrating nodules in the rectovaginal septum, uterosacral \nligaments, or bladder (Baușic et al. 2023). However, the sen-\nsitivity and specificity of ultrasound are highly dependent \non the operator's skill and may be limited in detecting peri-\ntoneal lesions or early-stage disease. According to the con-\nsensus statement by the International Deep Endometriosis \n\n Journal of Applied Genetics\nAnalysis (IDEA) group, systematic transvaginal ultrasound \nconducted by experienced examiners can reach high diag-\nnostic accuracy. Guerriero et al. (2016) report that for deep \nendometriosis, the sensitivity of TVUS ranges from 79 to \n98%, and specificity from 94 to 100%, depending on lesion \nlocation and scanning protocol (Bazot et al. 2004).\nMagnetic resonance imaging (MRI) has become a power-\nful complementary method. It’s excellent soft tissue contrast \nresolution and multiplanar imaging capabilities, which allow \nfor precise mapping of endometrial lesions, particularly in \ncomplex or deeply infiltrating cases. Studies have shown \nthat MRI sensitivity ranges from 77 to 93%, and specificity \nfrom 90 to 98%, depending on the imaging protocol and the \nradiologist's expertise (Fiaschetti et al. 2012). Functional \nMRI (fMRI), although primarily used in neuroscience, has \nrecently been applied to study altered brain connectivity in \nwomen with chronic pelvic pain and endometriosis (Szabo \net al. 2022).\nPreliminary studies have demonstrated abnormal activa-\ntion of pain-processing networks and altered connectivity in \nregions such as the anterior cingulate cortex and the insula. \nThese findings suggest that central sensitization may be \na hallmark of chronic endometriosis and could serve as a \npotential neuroimaging biomarker of disease severity and \npain chronicity (As-Sanie et al. 2016).\nTogether, these imaging and anthropometric biomark -\ners reflect both peripheral and central manifestations of \nendometriosis, offering a multifaceted approach that may \nimprove early diagnosis, patient risk stratification, and the \nunderstanding of disease mechanisms. Advanced imaging \ntechniques such as ultrasound, MRI, and more recently, \nfunctional MRI (fMRI), have been employed to visualize \nendometriotic lesions and assess their characteristics. Recent \nstudies have honed in on specific imaging features, patterns, \nand potential quantitative measures that can enhance the \naccuracy of diagnosis and characterization.\nMultimodal approaches and precision medicine\nThe diagnosis of endometriosis is not a one-size-fits-all \nendeavor. The interplay of various biomarkers, clinical \nparameters, and advanced technologies underscores the \nimportance of multimodal diagnostic approaches. Addition-\nally, the concept of precision medicine is emerging, where \nindividualized care is tailored based on a patient's unique \nbiomarker profile. While remarkable progress has been \nmade in identifying potential biomarkers for endometriosis, \nthe road to clinical implementation is ongoing. Emerging \ntechnologies, including artificial intelligence and machine \nlearning, promise to revolutionize diagnostic accuracy and \nefficiency.\nIn conclusion, these biomarkers offer a glimpse into the \ncomplexity of the disease, emphasizing the potential for \nenhanced diagnostic accuracy, individualized treatment, \nand improved patient outcomes. Ongoing research and col-\nlaborative efforts are vital in driving the field forward and \nharnessing the full potential of these biomarkers.\nThe most promising biomarkers \nfor endometriosis detection\nCA‑125: a widely studied marker\nCA-125, widely recognized as a tumor marker, has been \nextensively studied in the context of endometriosis (Feduniw \net al. 2024). Recent research continues to explore its role as \na diagnostic biomarker. While not specific to endometrio-\nsis, elevated CA-125 levels have shown promise in distin-\nguishing this condition from other gynecological disorders \n(Magalhães et al. 2021). This chapter focuses on the most \nrecent findings regarding CA-125 as a diagnostic tool.\nCA-125 is one of the most studied metabolic biomarkers \nin the context of non-invasive diagnosis of endometriosis. \nAlthough it lacks sufficient sensitivity to be used as a stan-\ndalone diagnostic test, it can serve as a useful complemen-\ntary tool, particularly in cases of moderate to severe disease. \nCA-125 is a membrane-bound glycoprotein often elevated \nin various pathological conditions, including ovarian cancer, \npelvic inflammatory disease, menstruation, and endometrio-\nsis. Its serum levels have been shown to correlate with dis-\nease stage and lesion burden in endometriosis.\nA comprehensive meta-analysis of 22 studies involving \n3626 women a threshold of 30 U/ml resulted in a sensitiv -\nity of 52 percent and specificity of 93 percent, indicating \nthat CA-125 may be more effective as a rule-in rather than \na rule-out diagnostic marker particularly in more advanced \nstages of the disease (Nisenblat et al. 2016). Sensitivity was \nsignificantly higher in moderate to severe endometriosis at \n63 percent compared to only 24 percent in minimal disease, \nsuggesting that lesion volume and location significantly \ninfluence circulating levels of this marker (Nisenblat et al. \n2016).\nA prospective cohort study confirmed similar findings \nshowing that a CA-125 threshold of 30 U/ml yielded a sen-\nsitivity of 57 percent and specificity of 96 percent with a \npositive likelihood ratio of 15.8 which provides substantial \nsupport for its diagnostic value in symptomatic patients \n(Szubert et al. 2012). Additionally, other studies have dem-\nonstrated that women with stage III and IV endometriosis \nhad mean CA-125 levels around 50 U/ml whereas women \nwithout endometriosis had mean levels closer to 7.8 U/ml \n(Somigliana et al. 2004).\nAlthough CA-125 levels can be influenced by menstrua-\ntion and other gynecological conditions its significant ele-\nvation in advanced endometriosis supports its inclusion in \n\nJournal of Applied Genetics \ndiagnostic panels for selected patients CA-125 may also be \nuseful in tracking treatment response or recurrence although \nits clinical utility in monitoring remains to be validated \nin larger longitudinal studies (Mol et al. 1998) despite its \nlimitations CA-125 continues to be the most accessible \nand widely used metabolic biomarker in endometriosis and \nongoing research aims to improve its specificity and sensi-\ntivity through combination with other markers or imaging \ntechniques (May et al. 2010).\nChen et al. (2021) examined differences in blood cells \nand tumor biomarkers between endometriosis patients and \ncontrols. They found notable discrepancies in blood cell \ncounts and tumor markers, with endometriosis patients hav-\ning altered levels of eosinophil, neutrophil count, and others. \nA diagnostic model using HGB, CA199, CA-125, and HE4 \nshowed a sensitivity of 85.4%, specificity of 78.83%, and an \nAUC of 0.900, suggesting enhanced diagnostic accuracy for \nearly endometriosis detection (Chen et al. 2021).\nHE4: a complementary marker\nHuman Epididymis Protein 4 (HE4) is a glycoprotein \ntraditionally associated with ovarian cancer diagnosis. \nRecent studies have explored its potential as a biomarker \nfor endometriosis, particularly in differentiating it from \nother gynecological conditions. For instance, a multicenter \nprospective study involving 981 patients assessed the util -\nity of HE4 in distinguishing endometriosis from adnexal \nmalignancies. The findings indicated that HE4 was positive \nin only 1.5% of endometriosis cases, compared to 64.6% \npositivity for CA125. This suggests that HE4 has high \nspecificity in excluding malignant disease in patients with \nendometriosis, especially when CA125 levels are elevated \n(Zapardiel et al. 2016).Another study evaluated the diag-\nnostic performance of HE4 compared to CA125 in differ -\nentiating between ovarian cancer and endometriosis. In \nthis sub-analysis, HE4 demonstrated superior performance \nwith an area under the curve (AUC) of 0.91, compared to \n0.81 for CA125. The Risk of Ovarian Malignancy Algo-\nrithm (ROMA), which incorporates both HE4 and CA125, \nachieved an AUC of 0.95, indicating enhanced diagnostic \naccuracy when combining these biomarkers (Braicu et al. \n2022). Furthermore, research has shown that HE4 levels are \nnot elevated in patients with endometriosis or benign ovar -\nian masses, whereas they are significantly higher in ovarian \ncancer patients. This reinforces the potential of HE4 as a \nspecific marker to distinguish between benign and malignant \novarian conditions (Anastasi et al. 2013).\nIn conclusion, HE4 holds promise as a specific marker \nfor distinguishing endometriosis from ovarian malignancies, \nespecially when used alongside CA125. The specificity of \nHE4 can improve diagnostic accuracy, especially in cases \nwhere CA125 is elevated (Abdalla et al. 2016).\nTGFBI and CA‑125\nTransforming Growth Factor Beta-Induced protein (TGFBI) \nis an extracellular matrix protein encoded by the TGFBI \ngene and regulated by TGF-β signaling. It plays a key role \nin cell adhesion, migration, and tissue remodeling. and is \ninvolved in several physiological and pathological processes \nincluding inflammation, fibrosis, and tumor progression. In \nthe context of endometriosis, elevated TGFBI levels may \nreflect abnormal tissue remodeling and immune activity \nassociated with the disease.\nA study by Janša et al. (2023) assessed the diagnostic \nvalue of TGFBI in combination with CA-125 for non-\ninvasive detection of endometriosis. (Janša et al. 2023 ). \nThe analysis was performed in two phases: discovery and \nvalidation. In the discovery cohort, (32 patients, 24 con-\ntrols), TGFBI levels were significantly higher in patients, \nwith a receiver operating characteristic (ROC) AUC of 0.77, \nsensitivity of 58%, and specificity of 84%. When combined \nwith CA-125 using a support vector machine (SVM) model, \ndiagnostic performance improved notably (AUC 0.91, sen-\nsitivity 88%, specificity 75%) (Janša et al. 2023). Validation \non a larger cohort (166 patients, 71 controls) confirmed the \nutility of the combined marker set, achieving an AUC of \n0.83, sensitivity of 83%, and specificity of 67%. CA-125 \nalone produced similar AUC (0.83) but with lower sensitiv-\nity (73%) and higher specificity (80%). Importantly, in early-\nstage disease (rASRM I–II), TGFBI outperformed CA-125 \n(AUC 0.74 vs. 0.63). For advanced stages (rASRM III–IV), \nthe combined model showed high diagnostic accuracy (AUC \n0.94, sensitivity 95%).(Janša et al. 2023). These findings \nhighlight TGFBI as a promising biomarker that enhances \ndiagnostic accuracy, particularly in early-stage endometrio-\nsis when used alongside CA-125. Further large-scale studies \nare warranted to validate these results and explore clinical \nintegration.\nInflammatory cytokines: insights into disease \nactivity\nRecent research has identified several interleukins and \nrelated cytokines as promising non-invasive biomarkers \nfor the diagnosis and monitoring of endometriosis. These \nimmune mediators, involved in regulating inflammation \nand tissue remodeling, often display altered expression pat-\nterns in patients with endometriosis compared to healthy \nindividuals.\nInterleukin-6 (IL-6): is a pro-inflammatory cytokine \nfrequently elevated in the serum and peritoneal fluid of \nwomen with endometriosis. A systematic review and \nmeta-analysis confirmed significantly higher IL-6 lev -\nels in affected individuals, highlighting its potential as a \n\n Journal of Applied Genetics\ndiagnostic biomarker. Furthermore, IL-6 has been pro -\nposed as a predictor of endometriosis-associated infertility \n(Krygere et al. 2024).\nInterleukin-1β (IL-1β): IL-1β, another pro-inflamma-\ntory cytokine, has been found at elevated levels in the \nserum of women with endometriosis. Research indicates \nthat IL-1β, along with IL-6 and TNF-α, could be used as \npredictors for endometriosis (Malutan et al. 2015).\nTumor Necrosis Factor-alpha (TNF-α): a key regulator \nof systemic inflammation, is also elevated in the serum \nof endometriosis patients. Its involvement in promoting \nangiogenesis and inflammatory signaling supports its util-\nity as a diagnostic indicator (Krygere et al. 2024).\nInterleukin-1 Receptor Antagonist (IL-1RA): is an anti-\ninflammatory cytokine that inhibits the activities of IL-1β. \nElevated levels of IL-1RA have been reported in the serum \nand peritoneal fluid of patients with endometriosis, indi-\ncating its involvement in the disease's pathophysiology \n(Werdel et al. 2024).\nInterleukin-10 (IL-10), another anti-inflammatory \ncytokine, suppresses immune responses and may contrib-\nute to immune tolerance mechanisms in endometriosis. \nElevated IL-10 levels have been reported in both serum \nand peritoneal fluid of affected individuals (Suen et al. \n2014).\nInterleukin-8 (IL-8): is a chemokine associated with \nchronic inflammation. Its expression is particularly increased \nduring the early stages of the disease and in patients with \nendometriomas, indicating its potential utility as a biomarker \nfor early detection (Arici et al. 1996; Cardoso et al. 2023).\nInterleukin-16 (IL-16): has been found at significantly \nhigher concentrations in the peritoneal fluid of patients with \nadvanced-stage endometriosis. This cytokine may play a role \nin initiating and sustaining peritoneal inflammation (Koga \net al. 2005; Krygere et al. 2024).\nInterleukin-18 (IL-18): involved in immune modulation, \nhas shown elevated levels in the peritoneal fluid of women \nwith minimal-to-mild endometriosis, although serum lev -\nels did not differ significantly. Further studies are needed \nto assess its diagnostic potential (Oku et al. 2004; Luo et al. \n2006).\nCombining multiple cytokines into a panel may enhance \ndiagnostic performance. For instance, a study combining \nIL-6, IL-8, TNF-α, high-sensitivity C-reactive protein, \nCA-125, and CA-19.9 achieved a sensitivity of 100% and \nspecificity of 84% for moderate-to-severe endometriosis \n(Krygere et al. 2024).\nLarge-scale, multicenter studies are essential to validate \nthese findings and to develop standardized, cytokine-based \nnon-invasive diagnostic tools. Measuring inflammatory \ncytokines such as IL-6, IL-8, and TNF-α not only offers \ndiagnostic value but also provides insights into disease activ-\nity and highlights potential therapeutic targets.\nMicroRNAs (miRNAs): Epigenetic biomarkers \nin endometriosis\nEpigenetic biomarkers, particularly microRNAs (miRNAs), \nhave gained substantial attention in the last few years. These \nsmall RNA molecules play a role in the regulation of gene \nexpression and have been found to exhibit distinct profiles \nin endometriosis. Research articles have identified specific \nmiRNA signatures associated with endometriosis, raising \nthe exciting prospect of non-invasive diagnostic tests based \non miRNA patterns. Furthermore, these biomarkers hold \nthe potential for predicting disease progression and thera-\npeutic response, paving the way for personalized treatment \nstrategies.\nIn a 2023 study, researchers explored the potential of a \nmicroRNA (miRNA) panel in combination with CA-125 as \na diagnostic biomarker for endometriosis. The aim was to \nassess how effectively this miRNA panel could detect early-\nstage endometriosis and distinguish it from more advanced \nstages of the disease. The study analyzed the serum levels \nof miR-199a, miR-122, miR-145, and miR-141 alongside \nCA-125, a well-established biomarker for endometriosis, in \nwomen diagnosed with endometriosis (early and advanced \nstages) compared to a control group of healthy women. \nThe team utilized quantitative polymerase chain reaction \n(qPCR) to profile miRNAs and enzyme-linked immunosorb-\nent assay (ELISA) to measure CA-125 levels. Diagnostic \naccuracy was evaluated based on sensitivity, specificity, and \nthe AUC from the receiver operating characteristic (ROC) \ncurve. The results showed that the combination of the four \nmiRNAs with CA-125 achieved excellent diagnostic perfor-\nmance, with 81.8% sensitivity, 92.6% specificity, and a 0.939 \nAUC—indicating strong diagnostic accuracy, especially for \ndistinguishing early-stage from advanced-stage endometrio-\nsis. The panel's ability to identify early stages of the disease \nis particularly important for timely diagnosis and interven-\ntion (Chen et al. 2023). The study concludes that combining \nmiRNA panels with CA-125 could significantly improve the \nnon-invasive diagnostic approach to endometriosis. While \nthe results are promising, further validation in larger, mul -\nticenter trials is needed to confirm these findings and assess \nthe clinical applicability of this biomarker panel in routine \nclinical practice.\nFuture prospects: the role of artificial intelligence \nand machine learning\nThis section concludes by exploring the potential of emerg-\ning technologies—particularly artificial intelligence (AI) and \nmachine learning (ML)—in analyzing complex biomarker \ndatasets to improve the diagnosis of endometriosis. Recent \nadvancements in these technologies have demonstrated \nsignificant promise in enhancing non-invasive diagnostic \n\nJournal of Applied Genetics \napproaches through sophisticated data integration and pat-\ntern recognition.\nA 2022 study introduced a non-invasive, blood-based \ndiagnostic model using AI to analyze the human miRNome, \nachieving 96.8% sensitivity, 100% specificity, and an AUC \nof 98.4%, These results suggest the potential of AI to replace \ninvasive diagnostic procedures such as laparoscopy (Ben-\ndifallah et al. 2024 ). Similarly, a 2024 study applied ML \nalgorithms to serum biomarkers, demonstrating that the \nrandom forest model combining CA125 and neutrophil-to-\nlymphocyte ratio (NLR) achieved an accuracy of 78.16%, \nsensitivity of 86.21%, and AUC of 0.85 (Zhao et al. 2024).\nIn another application, ML was used to identify 11 \nimmune-related genes regulated by 8 miRNAs as poten-\ntial diagnostic biomarkers, underlining the relevance of the \nimmune microenvironment in disease development (Tu et al. \n2024). Moreover, a 2022 study utilizing clinical and symp-\ntom-based features applied various ML algorithms to screen \nfor endometriosis, achieving AUCs between 0.91 and 0.95, \nwhich supports their potential as diagnostic aids in primary \ncare settings (Bendifallah et al. 2022).\nA comprehensive scoping review from the same year \nexamined 36 studies employing AI in endometriosis research \nand found that models such as logistic regression, deci-\nsion trees, random forests, and support vector machines \nhad pooled sensitivities ranging from 81.7% to 96.7% and \nspecificities between 70.7% and 91.6%, utilizing diverse \ndata inputs including clinical, imaging, and biomarker data \n(Sivajohan et al. 2022). Collectively, these findings collec-\ntively highlight the transformative potential of AI and ML \nin advancing non-invasive, accurate, and early diagnosis of \nendometriosis through the integration of biomarker data.\nNext-generation sequencing (NGS) has also contributed \nsignificantly to biomarker discovery. Multiple studies have \nidentified circulationg microRNAs in plasma with differ -\nential expression between women with laparoscopically \nconfirmed endometriosis and healthy controls, identifying \nup to 41 miRNAs as potential diagnostic markers (Nguyen \net al., 2020). Similarly, NGS combined with qRT-PCR has \nfurther validated these candidate miRNAs (Suryawanshi \net al., 2020). Broader genomic analyses using NGS have \nrevealed multiple genes implicated in endometriosis patho-\ngenesis, emphasizing its value in biomarker and therapeutic \ndiscovery (Ali et al., 2024). However, no clinically validated \ndiagnostic test based solely on NGS has yet been imple-\nmented, reflecting the translational gap between discovery \nand clinical application.\nIn contrast, AI and ML approaches have advanced closer \nto practical use. Systematic reviews indicate that AI mod-\nels trained on clinical, biomarker, imaging, and omics data \nachieve high diagnostic accuracy, with sensitivity ranging \nfrom 81.7% to 96.7% and specificity from 70.7% to 91.6%, \nusing methods such as logistic regression, random forest, \nand support vector machines (Tao et al., 2022). Projects like \nIMAGENDO demonstrated that AI-assisted integration of \nMRI and transvaginal ultrasound improved diagnostic accu-\nracy for detecting posterior cul-de-sac obliteration, raising \nthe AUC from approximately 65% to 90.6% (Avery et al. \n2024, IMAGENDO study). Deep learning architectures \nincluding Xception, Inception-V4, ResNet50, DenseNet, and \nEfficientNetB2 applied to ultrasound data achieved strong \ndiscriminatory power with AUC values of 0.85–0.90 (Chen \net al. 2023). Furthermore, explainable AI (XAI) approaches, \nsuch as U-Net with attention mechanisms and Grad-CAM, \nare being developed to enhance interpretability and clini -\ncal trustworthiness of AI predictions (Zhang et al. 2025).\nMultimodal frameworks like HAICOMM, integrating MRI \ndata with multi-rater expert assessments, have outperformed \nindividual radiologists in diagnostic classification (Wang \net al. 2024).\nA notable example of combining NGS and ML in a non-\ninvasive diagnostic is the saliva microRNA assay, commonly \nreferred to as Endotest. This assay sequences hundreds of \nsalivary miRNAs using NGS and applies an ML classifier, \nreporting very high diagnostic accuracy (AUC ≈0.98–0.99) \nin prospective case–control cohorts, including early and \nnon-ovarian disease. However, independent, real-world \nvalidation and reimbursement are still pending, and national \nhealth technology assessors have so far judged the evidence \npromising but insufficient for routine use (Bendifallah et al. \n2024). In summary, while NGS has advanced biomarker \ndiscovery, AI-based methods—particularly in imaging and \nmultimodal data integration—have reached a stage where \ndiagnostic performance rivals or exceeds expert evaluation. \nBoth approaches, however, require further validation, stand-\nardization, and regulatory approval before widespread clini-\ncal implementation.\nTo capture the current landscape of biomarker research \nin endometriosis, Table  2 provides a summary of the main \nbiomarker categories, representative examples, and their \npotential clinical applications.\nConclusions\nThe development of biomarkers for early detection of endo-\nmetriosis has made significant strides, particularly in the \nintegration of hormonal, genetic, epigenetic, immunologi-\ncal, metabolic, imaging, and multimodal approaches (Luo \net al. 2006; Anastasiu et al. 2020; Azeze et al. 2024) . These \nmarkers not only offer new insights into the pathophysiol-\nogy of the disease but also promise improved diagnostic \nprecision, allowing for earlier detection and personalized \ntreatment strategies.\nHormonal biomarkers, such as HE4 and CA-125, remain \ncentral to the diagnostic process, with HE4 showing promise \n\n Journal of Applied Genetics\nTable 2  Summary of major biomarker categories studied in endometriosis and their potential clinical applications\nBiomarker Type Example(s) Sample Source Diagnostic Role/Comments Ref\nClassical serum markers CA-125, HE4, TGFBI Serum (blood) Moderate sensitivity and specificity; \nbest used in combination with other \nbiomarkers\nCA-125 already used clinically (support-\nive, not definitive), HE4, TGFBI-\nUnder investigation, not yet routine\n(Zhao et al. 2024)\n(Feduniw et al. 2024)\n(Zapardiel et al. 2016)\n(Braicu et al. 2022)\n(Janša et al. 2023)\nInflammatory cytokines IL-6, IL-8, TNF-α, IL-10. IL-16, IL-18 Serum (blood) or peritoneal fluid Reflects local and systemic inflammation; \ncurrently under active investigation for \ndiagnostic use. Under investigation\n(Krygere et al. 2024)\n(Malutan et al. 2015)\n(Suen et al. 2014)\n(Cardoso et al. 2023)\n(Q. Luo et al. 2006; Oku et al. 2004)\nHormonal markers Estrogen, Progesteron, LH, FSH Serum (blood) Altered hormonal profiles observed; \npotential supportive markers in specific \nclinical contexts\nUsed clinically (hormonal profiling, sup-\nportive,\nnot diagnostic-specific)\n(Gao et al. 2019)\n(Gjorgoska & Rizner 2024; \nMcGrath et al., 2023; B. Zhao \net al. 2022)\nMicroRNAs miR-17-5p, miR-451a, let-7b-5p Serum, plasma, tissue, saliva Emerging as promising non-invasive \nbiomarkers; with potential for early and \naccurate diagnosis\nUnder investigation\n(Hon et al. 2024)\n(Vanhie et al. 2024)\nEpigenetic markers DNA methylation profiles, mQTLs Blood or tissue DNA/RNA Help identify disease susceptibility and \nprogression; still largely in the explora-\ntory research phase. Under investigation\n(Bendifallah et al. 2024)\n(Mortlock et al. 2023)\n(Ducreux et al. 2025)\n(Ortiz et al. 2021)\nGenetic markers Variants in CUX2, CLMP, CEP131, etc Blood or tissue DNA Associated with disease risk; may inform \npersonalized therapeutic strategies in \nthe future\nUnder investigation\n(Kucukakcali et al., 2025)\nImmune markers Regulatory T cell levels, cytokine profiles Serum (blood) or tissue samples Reflect immune dysregulation; potential \nrole in disease staging and monitoring. \nUnder investigation\n(Tarokh et al. 2019; Pashizeh et al. \n2020; Olkowska-Truchanowicz \net al. 2021)\nMetabolites Altered metabolic signatures Serum (blood) or urine Under active study for diagnostic panels; \nstrong potential for non-invasive test-\ning. Under investigation\n(Dutta et al. 2018)\n(Li et al. 2018)\nMultimodal approaches Combined biomarker panels, AI-\nenhanced analysis\n(e.g., Endotest saliva miRNA test, \nIMAGENDO MRI/TVUS integration, \nRandom forest with CA125 + NLR)\nBlood, imaging, multi-omics Improve diagnostic sensitivity and speci-\nficity; represent the future of precision \ndiagnosis. Under investigation\n(Krygere et al. 2024)\n(Bendifallah et al. 2024)\n(Sivajohan et al. 2022)\n\nJournal of Applied Genetics \nin distinguishing endometriosis from other gynecological \nconditions (Anastasi et al. 2013; Braicu et al. 2022). The \ncombination of these markers significantly enhances diag-\nnostic sensitivity and specificity, particularly in the context \nof early-stage disease.\nTGFBI, a marker of tissue fibrosis and extracellular \nmatrix remodeling, shows potential in the monitoring of dis-\nease progression (Janša et al. 2023). It can serve as a useful \nbiomarker to assess disease activity, aiding in the prediction \nof endometriosis recurrence and the evaluation of therapeu-\ntic interventions.\nInflammatory cytokines, particularly IL-6, IL-1β, TNF-\nα, and other interleukins, play a crucial role in the inflam-\nmatory response associated with endometriosis (Oală et al. \n2024; Krygere et al. 2024). The levels of these cytokines \ncorrelate with disease activity and can be used to monitor \nthe inflammatory status, providing valuable information for \ntreatment strategies.\nMiRNA profiles represent an evolving approach in bio-\nmarker diagnostics. Changes in miRNA expression pat-\nterns, especially in circulating samples, offer insights into \nthe molecular mechanisms underlying endometriosis and its \nprogression (Wang et al. 2013; Leonova et al. 2021). The \nability to detect these miRNAs non-invasively could signifi-\ncantly improve early diagnosis, particularly when used in \nconjunction with other biomarkers such as CA-125.\nThe 2D:4D ratio, a measure of prenatal hormone expo-\nsure, is a promising hormonal biomarker that may help iden-\ntify individuals at higher risk of developing endometriosis \n(Buggio et al. 2023). While the 2D:4D ratio warrants further \nexploration, it offers a potential early indicator for personal-\nized risk assessment.\nThe AGD has gained attention for its role in the develop-\nment and progression of endometriosis. Changes in AGD \nmay offer insights into the prenatal hormonal environment \nand the molecular mechanisms underlying the disease, \npotentially assisting in the identification of patients at higher \nrisk of severe disease progression (Zamani et al. 2023).\nBioimaging techniques are emerging as vital tools in the \nnon-invasive diagnosis and monitoring of endometriosis. \nAdvances in MRI, ultrasound, and other imaging modali-\nties allow for the visualization of endometriotic lesions and \nprovide real-time monitoring of disease progression (Avery \net al. 2024). These techniques are integral in improving \nthe accuracy of endometriosis diagnosis, especially when \ncombined with biomarker analysis, offering a more holistic \napproach to patient care.\nEpigenetic markers, including DNA methylation patterns, \nhistone modifications, and non-coding RNAs, offer valuable \ninsights into the molecular changes that underpin the patho-\ngenesis of endometriosis (Luo et al. 2024). These markers \nnot only enhance the understanding of the disease but also \nprovide the potential for identifying new therapeutic targets. \nThis table summarizes the main classes of biomarkers explored for endometriosis detection, including examples, sample sources, and diagnostic relevance. The integration of classical mark -\ners, inflammatory profiles, hormonal alterations, microRNAs, genetic and epigenetic factors, immune signatures, and metabolic changes reflects the complexity of the disease. Multimodal \napproaches combining multiple biomarkers with artificial intelligence represent the future direction of improving diagnostic precision.\nTable 2  (continued)\nBiomarker Type Example(s) Sample Source Diagnostic Role/Comments Ref\nImaging\n(Neuroimaging)\nMRI, fMRI,\nTransvaginal ultrasound\nDetect ovarian and deep lesions, assess \nCNS changes\nCombine imaging with metabolic bio-\nmarkers to improve diagnostic accuracy\nfMRI- evaluates central sensitization; \nassesses activation of brain pain net-\nworks\nAlready used clinically (imaging stand-\nards in diagnosis)\n(Borghese et al. 2015)\n(Fiaschetti et al. 2012)\n(Baușic et al. 2023)\nPhenotypic 2D:4D ratio, birth weight,\nAGD\nEarly identification of at-risk individuals Merge phenotypic traits with imaging to \nimprove patient stratification and risk \nprofiling\nUnder investigation (phenotypic/epide-\nmiological\nrisk markers\n(Buggio et al. 2023)\n(Zamani et al. 2023)\n(Borghese et al. 2015)\n\n Journal of Applied Genetics\nEpigenetic modifications are also crucial for understanding \nthe transgenerational impact of the disease, particularly in \nthe context of maternal transmission of endometriosis risk.\nFurthermore, multimodal approaches that integrate \nbioimaging with molecular biomarkers, such as miRNAs, \ncytokines, and genetic markers, represent the future of endo-\nmetriosis diagnostics. The application of artificial intelli-\ngence (AI) to analyze complex datasets can significantly \nenhance diagnostic accuracy, enabling personalized medi-\ncine strategies and improving patient outcomes.\nFuture directions in clinical implementation\nThe early and accurate detection of endometriosis is para -\nmount, particularly in avoiding the long-term complications \nassociated with the disease, such as infertility. One of the \nmost significant challenges in current clinical practice is \nthe reliance on invasive diagnostic techniques, which often \nresult in delayed diagnoses and the potential for irreversible \ndamage. Therefore, there is an urgent need to identify and \nimplement non-invasive biomarkers that can provide quick, \nreliable, and minimally disruptive diagnostics.\nBy incorporating hormonal, genetic, epigenetic, inflam-\nmatory, and imaging biomarkers, we move towards the \npossibility of diagnosing endometriosis in its early stages, \nbefore it has the opportunity to cause extensive damage to \nthe reproductive organs. The use of bioimaging technologies \ncombined with circulating biomarkers, such as miRNAs, \ninflammatory cytokines, and protein markers, presents a \npromising non-invasive approach for detecting endometrio-\nsis. This not only improves the speed of diagnosis but also \nreduces the need for invasive surgeries that carry risk and \ncan lead to delays in treatment. Implementing non-invasive \ndiagnostic methods would revolutionize clinical manage-\nment, enabling earlier intervention that could prevent the \ndisease from causing significant tissue damage, reducing the \nrisk of infertility, and improving overall patient outcomes. \nWith such methods, the focus could shift from reactive, inva-\nsive procedures to proactive, preventive care, empowering \nclinicians to manage endometriosis more effectively and pro-\nviding patients with the opportunity for timely, personalized \ntreatment.\nThus, advancing the search for non-invasive diagnostic \nsolutions is essential in transforming the approach to endo-\nmetriosis care, offering a way to diagnose this debilitating \ndisease early and reduce its impact on reproductive health.\nAuthor contribution Michalina Kliber: conceptualization, writ-\ning—original draft, review and editing, literature searching; Klaudia \nKulczynska-Figurny: writing-review and editing, figure prepara-\ntion; Andrzej Pławski editing, supervision; Paweł Piotr Jagodziński \nsupervision.\nData Availability No new data were generated or analyzed in this study. \nAll information presented is based on previously published studies, \nwhich are appropriately cited in the manuscript.\nDeclarations \nEthics approval This article does not contain any studies with human \nparticipants or animals performed by any of the authors.\nConsent to participate Not applicable/not required.\nConsent for publication Not applicable/not required.\nCompeting interest The authors declare that there is no conflicts of \ninterest.\nOpen Access This article is licensed under a Creative Commons Attri-\nbution 4.0 International License, which permits use, sharing, adapta-\ntion, distribution and reproduction in any medium or format, as long \nas you give appropriate credit to the original author(s) and the source, \nprovide a link to the Creative Commons licence, and indicate if changes \nwere made. The images or other third party material in this article are \nincluded in the article’s Creative Commons licence, unless indicated \notherwise in a credit line to the material. If material is not included in \nthe article’s Creative Commons licence and your intended use is not \npermitted by statutory regulation or exceeds the permitted use, you will \nneed to obtain permission directly from the copyright holder. To view a \ncopy of this licence, visit http://creativecommons.org/licenses/by/4.0/.\nReferences\nAbdalla N, Pazura M, Słomka A et al (2016) The role of HE4 and \nCA125 in differentiation between malignant and non-malignant \nendometrial pathologies. Ginekol Pol 87:781–786. https:// doi.  \norg/ 10. 5603/ GP. 2016. 0088\nAlAshqar A, Reschke L, Kirschen GW, Borahay MA (2021) Role of \ninflammation in benign gynecologic disorders: from pathogen-\nesis to novel therapies†. Biol Reprod 105:7–31. https:// doi. org/ \n10. 1093/ biolre/ ioab0 54\nAlbertsen HM, Chettier R, Farrington P, Ward K (2013) Genome-wide \nassociation study link novel loci to endometriosis. PLoS ONE \n8:e58257. https:// doi. org/ 10. 1371/ journ al. pone. 005825\nAmanda CR, Asmarinah, Hestiantoro A et al (2024) Gene expres-\nsion of aromatase, SF-1, and HSD17B2 in menstrual blood as \nnoninvasive diagnostic biomarkers for endometriosis. European \nJ Obstet Gynecol Reprod Biol 301:95–101. https:// doi. org/ 10. \n1016/j. ejogrb. 2024. 07. 061\nAnastasi E, Granato T, Falzarano R et al (2013) The use of HE4, \nCA125 and CA72-4 biomarkers for differential diagnosis \nbetween ovarian endometrioma and epithelial ovarian cancer. J \nOvarian Res 6:44. https:// doi. org/ 10. 1186/ 1757- 2215-6- 44\nAnastasiu CV, Moga MA, Elena Neculau A et al (2020) Biomarkers \nfor the noninvasive diagnosis of endometriosis: state of the art \nand future perspectives. Int J Mol Sci 21:1750. https:// doi. org/ \n10. 3390/ ijms2 10517 50\nArici A, Tazuke SI, Attar E et al (1996) Interleukin-8 concentration in \nperitoneal fluid of patients with endometriosis and modulation \nof interleukin-8 expression in human mesothelial cells. Mol Hum \nReprod 2:40–45. https:// doi. org/ 10. 1093/ molehr/ 2.1. 40\nAs-Sanie S, Kim J, Schmidt-Wilcke T et al (2016) Functional connec-\ntivity is associated with altered brain chemistry in women with \nendometriosis-associated chronic pelvic pain. J Pain 17:1–13. \nhttps:// doi. org/ 10. 1016/j. jpain. 2015. 09. 008\n\nJournal of Applied Genetics \nAvery JC, Deslandes A, Freger SM et al (2024) Noninvasive diagnostic \nimaging for endometriosis part 1: a systematic review of recent \ndevelopments in ultrasound, combination imaging, and artificial \nintelligence. Fertil Steril 121:164–188. https:// doi. org/ 10. 1016/j. \nfertn stert. 2023. 12. 008\nAzeze GG, Wu L, Alemu BK et al (2024) Proteomics approach to \ndiscovering non-invasive diagnostic biomarkers and understand-\ning the pathogenesis of endometriosis: a systematic review and \nmeta-analysis. J Transl Med 22:685. https:// doi. org/ 10. 1186/  \ns12967- 024- 05474-3\nBaușic AIG, Matasariu DR, Manu A, Brătilă E (2023) Transvagi-\nnal ultrasound vs. magnetic resonance imaging: What is the \noptimal imaging modality for the diagnosis of endometriosis? \nBiomedicines 11:2609. https:// doi. org/ 10. 3390/ biome dicin  \nes111 02609\nBazot M, Thomassin I, Hourani R et al (2004) Diagnostic accuracy \nof transvaginal sonography for deep pelvic endometriosis. \nUltrasound Obstet Gynecol 24:180–185. https:// doi. org/ 10. \n1002/ uog. 1108\nBekbolatova M, Mayer J, Ong CW, Toma M (2024) Transforma -\ntive potential of AI in healthcare: definitions, applications, \nand navigating the ethical landscape and public perspectives. \nHealthcare (Basel) 12:125. https:// doi. org/ 10. 3390/ healt hcare \n12020 125\nBendifallah S, Puchar A, Suisse S et al (2022) Machine learning algo-\nrithms as new screening approach for patients with endometrio-\nsis. Sci Rep 12:639. https:// doi. org/ 10. 1038/ s41598- 021- 04637-2\nBendifallah S, Dabi Y, Suisse S et al (2024) Saliva-based microrna \ndiagnostic signature for the superficial peritoneal endometriosis \nphenotype. European J Obstet Gynecol Reprod Biol 297:187–\n196. https:// doi. org/ 10. 1016/j. ejogrb. 2024. 04. 020\nBlanco LP, Salmeri N, Temkin SM et al (2025) Endometriosis and \nAutoimmunity. Autoimmun Rev 24:103752. https:// doi. org/ 10. \n1016/j. autrev. 2025. 103752\nBorghese B, Sibiude J, Santulli P et al (2015) Low birth weight is \nstrongly associated with the risk of deep infiltrating endometrio-\nsis: results of a 743 case-control study. PLoS ONE 10:e0117387. \nhttps:// doi. org/ 10. 1371/ journ al. pone. 01173 87\nBraicu EI, Krause CL, Torsten U et al (2022) HE4 as a serum bio-\nmarker for the diagnosis of pelvic masses: a prospective, mul-\nticenter study in 965 patients. BMC Cancer 22:831. https:// doi.  \norg/ 10. 1186/ s12885- 022- 09887-5\nBuggio L, Reschini M, Viganò P et al (2023) Is there a correlation \nbetween the second-to-four digit ratio (2D:4D) and endometrio-\nsis? Results of a case-control study. J Clin Med 12:2040. https:// \ndoi. org/ 10. 3390/ jcm12 052040\nBulun SE, Yilmaz BD, Sison C et al (2019) Endometriosis. Endocr Rev \n40:1048–1079. https:// doi. org/ 10. 1210/ er. 2018- 00242\nCao W-G, Morin M, Metz C et al (2005) Stimulation of macrophage \nmigration inhibitory factor expression in endometrial stromal \ncells by interleukin 1, beta involving the nuclear transcription \nfactor NFkappaB. Biol Reprod 73:565–570. https:// doi. org/ 10. \n1095/ biolr eprod. 104. 038331\nCardoso JV, Machado DE, da Silva MC et al (2023) Influence of inter-\nleukin-8 polymorphism on endometriosis-related pelvic pain. \nHum Immunol 84:561–566. https:// doi. org/ 10. 1016/j. humimm. \n2023. 07. 004\nCarli C, Metz CN, Al-Abed Y et al (2009) Up-regulation of cyclooxy -\ngenase-2 expression and prostaglandin E2 production in human \nendometriotic cells by macrophage migration inhibitory factor: \ninvolvement of novel kinase signaling pathways. Endocrinology \n150:3128–3137. https:// doi. org/ 10. 1210/ en. 2008- 1088\nChen T, Wei J-L, Leng T et al (2021) The diagnostic value of the \ncombination of hemoglobin, CA199, CA125, and HE4 in endo-\nmetriosis. J Clin Lab Anal 35:e23947. https:// doi. org/ 10. 1002/ \njcla. 23947\nChen G, Guo J, Li W et al (2023) Diagnostic value of the combina-\ntion of circulating serum miRNAs and CA125 in endometriosis. \nMedicine (Baltimore) 102:e36339. https:// doi. org/ 10. 1097/ MD. \n00000 00000 036339\nCrestani A, Arfi A, Ploteau S et al (2020) Anogenital distance in adult \nwomen is a strong marker of endometriosis: results of a prospec-\ntive study with laparoscopic and histological findings. Human \nReprod Open 2020:hoaa023. https:// doi. org/ 10. 1093/ hropen/ \nhoaa0 23\nde Magalhães JS, Jammal MP, Crispim PCA et al (2021) Role of \nbiomarkers CA-125, CA-15.3 and CA-19.9 in the distinction \nbetween endometriomas and ovarian neoplasms. Biomarkers \n26:268–274. https:// doi. org/ 10. 1080/ 13547 50X. 2021. 18854 90\nDucreux B, Patrat C, Firmin J et al (2025) Systematic review on the \nDNA methylation role in endometriosis: current evidence and \nperspectives. Clin Epigenetics 17:32. https:// doi. org/ 10. 1186/ \ns13148- 025- 01828-w\nDungate B, Tucker DR, Goodwin E, Yong PJ (2024) Assessing the \nutility of artificial intelligence in endometriosis: Promises and \npitfalls. Womens Health (Lond) 20:17455057241248120. https:// \ndoi. org/ 10. 1177/ 17455 05724 12481 21\nDutta M, Singh B, Joshi M et al (2018) Metabolomics reveals per -\nturbations in endometrium and serum of minimal and mild \nendometriosis. Sci Rep 8:6466. https:// doi. org/ 10. 1038/  \ns41598- 018- 23954-7\nEncalada Soto D, Rassier S, Green IC et al (2022) Endometriosis bio-\nmarkers of the disease: an update. Curr Opin Obstet Gynecol \n34:210–219. https:// doi. org/ 10. 1097/ GCO. 00000 00000 000798\nFeduniw S, Pruc M, Ciebiera M et al (2024) Current evidence on \nCA-125 levels in differentiation between endometriomas and \nendometriosis-associated ovarian cancer – a systematic review \nand meta-analysis. J Endometr Pelvic Pain Disord 16:154–159. \nhttps:// doi. org/ 10. 1177/ 22840 26524 12753 49\nFiaschetti V, Crusco S, Meschini A et al (2012) Deeply infiltrating \nendometriosis: evaluation of retro-cervical space on MRI after \nvaginal opacification. Eur J Radiol 81:3638–3645. https:// doi.  \norg/ 10. 1016/j. ejrad. 2011. 06. 058\nGajbhiye R, Bendigeri T, Ghuge A et al (2017) Panel of autoimmune \nmarkers for noninvasive diagnosis of minimal-mild endometrio-\nsis. Reprod Sci 24:413–420. https:// doi. org/ 10. 1177/ 19337 19116 \n657190\nGao Y, Shen M, Ma X et al (2019) Seven hormonal biomarkers for \ndiagnosing endometriosis: meta-analysis and adjusted indi-\nrect comparison of diagnostic test accuracy. J Minim Invasive \nGynecol 26:1026-1035.e4. https:// doi. org/ 10. 1016/j. jmig. 2019. \n04. 004\nGjorgoska M, Rizner TL (2024) The effect of androgens on the risk \nof endometriosis sub-phenotypes and ovarian neoplasms: a \nMendelian randomization study. J Steroid Biochem Mol Biol \n239:106482. https:// doi. org/ 10. 1016/j. jsbmb. 2024. 106482\nHanna MG, Pantanowitz L, Dash R, et al (2025) Future of artificial \nintelligence—machine learning trends in pathology and medi-\ncine. Modern Pathol 38. https:// doi. org/ 10. 1016/j. modpat. 2025. \n100705\nHarada T, Yoshioka H, Yoshida S et al (1997) Increased interleukin-6 \nlevels in peritoneal fluid of infertile patients with active endo-\nmetriosis. Am J Obstet Gynecol 176:593–597. https:// doi. org/ 10. \n1016/ s0002- 9378(97) 70553-2\nHarden S, Tan TY, Ku CW et al (2023) Peritoneal autoantibody profil-\ning identifies p53 as an autoantibody target in endometriosis. \nFertil Steril 120:176–187. https:// doi. org/ 10. 1016/j. fertn stert. \n2023. 02. 025\nHo HN, Wu MY, Chao KH et al (1996) Decrease in interferon gamma \nproduction and impairment of T-lymphocyte proliferation in peri-\ntoneal fluid of women with endometriosis. Am J Obstet Gynecol \n175:1236–1241. https:// doi. org/ 10. 1016/ s0002- 9378(96) 70034-0\n\n Journal of Applied Genetics\nHon J-X, Wahab NA, Karim AKA et al (2024) Exploring the role of \nMicroRNAs in progesterone and estrogen receptor expression in \nendometriosis. Biomedicines 12:2218. https:// doi. org/ 10. 3390/ \nbiome dicin es121 02218\nHou S, Xu H, Lei S, Zhao D (2024) Overexpressed nicotinamide \nN-methyltransferase in endometrial stromal cells induced by \nmacrophages and estradiol contributes to cell proliferation in \nendometriosis. Cell Death Discovery 10:463. https:// doi. org/ 10. \n1038/ s41420- 024- 02229-3\nHuang L, Chen M, Long L et al (2021) Low basal serum testosterone \nlevel is detrimental to the embryo implantation in the patients \nwith severe endometriosis. J Obstet Gynecol Res 47:2166–2174. \nhttps:// doi. org/ 10. 1111/ jog. 14791\nJanša V, Pušić Novak M, Ban Frangež H, Rižner TL (2023) TGFBI as \na candidate biomarker for non-invasive diagnosis of early-stage \nendometriosis. Hum Reprod 38:1284–1296. https:// doi. org/ 10. \n1093/ humrep/ dead0 91\nJeung I, Cheon K, Kim M-R (2016) Decreased cytotoxicity of periph-\neral and peritoneal natural killer cell in endometriosis. Biomed \nRes Int 2016:2916070. https:// doi. org/ 10. 1155/ 2016/ 29160 70\nJoshi NR, Miyadahira EH, Afshar Y et al (2017) Progesterone resist-\nance in endometriosis is modulated by the altered expression of \nmicroRNA-29c and FKBP4. J Clin Endocrinol Metab 102:141–\n149. https:// doi. org/ 10. 1210/ jc. 2016- 2076\nKoga K, Osuga Y, Yoshino O et al (2005) Elevated interleukin-16 lev-\nels in the peritoneal fluid of women with endometriosis may be a \nmechanism for inflammatory reactions associated with endome-\ntriosis. Fertil Steril 83:878–882. https:// doi. org/ 10. 1016/j. fertn \nstert. 2004. 12. 004\nKrygere L, Jukna P, Jariene K, Drejeriene E (2024) Diagnostic poten-\ntial of cytokine biomarkers in endometriosis: challenges and \ninsights. Biomedicines 12:2867. https:// doi. org/ 10. 3390/ biome \ndicin es121 22867\nLalami I, Abo C, Borghese B et al (2021) Genomics of endometriosis: \nfrom genome wide association studies to exome sequencing. Int \nJ Mol Sci 22:7297. https:// doi. org/ 10. 3390/ ijms2 21472 97\nLeonova A, Turpin VE, Agarwal SK et al (2021) A critical appraisal \nof the circulating levels of differentially expressed microRNA in \nendometriosis†. Biol Reprod 105:1075–1085. https:// doi. org/ 10. \n1093/ biolre/ ioab1 34\nLi J, Guan L, Zhang H et al (2018) Endometrium metabolomic profil-\ning reveals potential biomarkers for diagnosis of endometriosis \nat minimal-mild stages. Reprod Biol Endocrinol 16:42. https://  \ndoi. org/ 10. 1186/ s12958- 018- 0360-z\nLukac S, Schmid M, Pfister K et al (2022) Extragenital endometrio-\nsis in the differential diagnosis of non- gynecological diseases. \nDeutsches Arzteblatt Int 119:361–367. https:// doi. org/ 10. 3238/ \narzte bl. m2022. 0176\nLuo Q, Ning W, Wu Y et al (2006) Altered expression of interleu-\nkin-18 in the ectopic and eutopic endometrium of women with \nendometriosis. J Reprod Immunol 72:108–117. https:// doi. org/ \n10. 1016/j. jri. 2006. 03. 003\nLuo L, Zhao L, Cui L et al (2024) The roles of chromatin regulatory \nfactors in endometriosis. J Assist Reprod Genet 41:863–873. \nhttps:// doi. org/ 10. 1007/ s10815- 024- 03026-8\nMalutan AM, Drugan T, Costin N et  al (2015) Clinical \nimmunology<br>Pro-inflammatory cytokines for evaluation of \ninflammatory status in endometriosis. Central European J Immu-\nnol 40:96–102. https:// doi. org/ 10. 5114/ ceji. 2015. 50840\nMay KE, Conduit-Hulbert SA, Villar J et al (2010) Peripheral biomark-\ners of endometriosis: a systematic review. Hum Reprod Update \n16:651–674. https:// doi. org/ 10. 1093/ humupd/ dmq009\nMay KE, Villar J, Kirtley S et al (2011) Endometrial alterations in \nendometriosis: a systematic review of putative biomarkers. Hum \nReprod Update 17:637–653. https://  doi. org/ 10. 1093/ humupd/ \ndmr013\nMcGrath IM, International Endometriosis Genetics Consortium, Mont-\ngomery GW, Mortlock S (2023) Polygenic risk score phenome-\nwide association study reveals an association between endome-\ntriosis and testosterone. BMC Med 21:482. https:// doi. org/ 10. \n1186/ s12916- 023- 03184-z\nMendiola J, Roca M, Mínguez-Alarcón L et al (2012) Anogenital dis-\ntance is related to ovarian follicular number in young Spanish \nwomen: a cross-sectional study. Environ Health 11:90. https://  \ndoi. org/ 10. 1186/ 1476- 069X- 11- 90\nMol BW, Bayram N, Lijmer JG et al (1998) The performance of \nCA-125 measurement in the detection of endometriosis: a meta-\nanalysis. Fertil Steril 70:1101–1108. https:// doi. org/ 10. 1016/  \ns0015- 0282(98) 00355-0\nMortlock S, Houshdaran S, Kosti I et al (2023) Global endometrial \nDNA methylation analysis reveals insights into mQTL regula-\ntion and associated endometriosis disease risk and endome-\ntrial function. Commun Biol 6:780. https:// doi. org/ 10. 1038/  \ns42003- 023- 05070-z\nMurphy AA (2002) Clinical aspects of endometriosis. Annals New \nYork Acad Sci 955:1–10. https:// doi. org/ 10. 1111/j. 1749- 6632. \n2002. tb027 60.x\nNisenblat V, Bossuyt PMM, Shaikh R et al (2016) Blood biomark -\ners for the non-invasive diagnosis of endometriosis. Cochrane \nDatabase Syst Rev 2016:CD012179. https:// doi. org/ 10. 1002/  \n14651 858. CD012 179\nOală IE, Mitranovici M-I, Chiorean DM et al (2024) Endometrio-\nsis and the role of pro-inflammatory and anti-inflammatory \ncytokines in pathophysiology: a narrative review of the lit-\nerature. Diagnostics (Basel) 14:312. https:// doi. org/ 10. 3390/  \ndiagn ostic s1403 0312\nOku H, Tsuji Y, Kashiwamura S-I et al (2004) Role of IL-18 in patho-\ngenesis of endometriosis. Hum Reprod 19:709–714. https:// doi. \norg/ 10. 1093/ humrep/ deh108\nOlkowska-Truchanowicz J, Białoszewska A, Zwierzchowska A et al \n(2021) Peritoneal fluid from patients with ovarian endometrio-\nsis displays immunosuppressive potential and stimulates Th2 \nresponse. Int J Mol Sci 22:8134. https:// doi. org/ 10. 3390/ ijms2 \n21581 34\nOrtiz CN, Torres-Reverón A, Appleyard CB (2021) Metabolomics \nin endometriosis: challenges and perspectives for future stud-\nies. Reprod Fertility 2:R35–R50. https:// doi. org/ 10. 1530/  \nRAF- 20- 0047\nOthman ER, Markeb AA, Khashbah MY et al (2021) Markers of local \nand systemic estrogen metabolism in endometriosis. Reprod Sci \n28:1001–1011. https:// doi. org/ 10. 1007/ s43032- 020- 00383-4\nOverton C, Fernandez-Shaw S, Hicks B et al (1996) Peritoneal fluid \ncytokines and the relationship with endometriosis and pain. Hum \nReprod 11:380–386. https:// doi. org/ 10. 1093/ humrep/ 11.2. 380\nPashizeh F, Mansouri R, Davari-Tanha F et al (2020) Alterations of \nCD4+T cell subsets in blood and peritoneal fluid in different \nstages of endometriosis. Int J Fertil Steril 14:201–208. https://  \ndoi. org/ 10. 22074/ ijfs. 2020. 6127\nPeters HE, Laeven CHC, Trimbos CJMA et al (2020) Anthropometric \nbiomarkers for abnormal prenatal reproductive hormone expo-\nsure in women with Mayer-Rokitanksy-Küster-Hauser syndrome, \npolycystic ovary syndrome, and endometriosis. Fertil Steril \n114:1297–1305. https:// doi. org/ 10. 1016/j. fertn stert. 2020. 06. 029\nPunnonen J, Teisala K, Ranta H et al (1996) Increased levels of inter -\nleukin-6 and interleukin-10 in the peritoneal fluid of patients with \nendometriosis. Am J Obstet Gynecol 174:1522–1526. https:// doi. \norg/ 10. 1016/ s0002- 9378(96) 70600-2\nRafi U, Ahmad S, Bokhari SS et al (2021) Association of inflamma-\ntory markers/cytokines with cardiovascular risk manifestation in \npatients with endometriosis. Mediators Inflamm 2021:3425560. \nhttps:// doi. org/ 10. 1155/ 2021/ 34255 60\n\nJournal of Applied Genetics \nRahal D, Andrade F, Nisihara R (2021) Insights into the role of com-\nplement system in the pathophysiology of endometriosis. Immu-\nnol Lett 231:43–48. https:// doi. org/ 10. 1016/j. imlet. 2021. 01. 005\nRavaggi A, Bergamaschi C, Galbiati C et al (2024) Circulating serum \nmicro-RNA as non-invasive diagnostic biomarkers of endome-\ntriosis. Biomedicines 12:2393. https:// doi. org/ 10. 3390/ biome \ndicin es121 02393\nRiccio LGC, Baracat EC, Chapron C et al (2017) The role of the B \nlymphocytes in endometriosis: a systematic review. J Reprod \nImmunol 123:29–34. https:// doi. org/ 10. 1016/j. jri. 2017. 09. 001\nSánchez-Ferrer ML, Mendiola J, Jiménez-Velázquez R et al (2017) \nInvestigation of anogenital distance as a diagnostic tool in endo-\nmetriosis. Reprod Biomed Online 34:375–382. https:// doi. org/  \n10. 1016/j. rbmo. 2017. 01. 002\nSapkota Y, Steinthorsdottir V, Morris AP et al (2017) Meta-analysis \nidentifies five novel loci associated with endometriosis highlight-\ning key genes involved in hormone metabolism. Nat Commun \n8:15539. https:// doi. org/ 10. 1038/ ncomm s15539\nSimko S, Wright KN (2022) The future of diagnostic laparoscopy – \ncons in reproduction and fertility 3:R91–R95. https:// doi. org/ 10. \n1530/ RAF- 22- 0007\nSimoens S, Dunselman G, Dirksen C et al (2012) The burden of endo-\nmetriosis: costs and quality of life of women with endometrio-\nsis and treated in referral centres. Hum Reprod 27:1292–1299. \nhttps:// doi. org/ 10. 1093/ humrep/ des073\nSivajohan B, Elgendi M, Menon C et al (2022) Clinical use of artificial \nintelligence in endometriosis: a scoping review. NPJ Digit Med \n5:109. https:// doi. org/ 10. 1038/ s41746- 022- 00638-1\nSomigliana E, Viganò P, Tirelli AS et al (2004) Use of the concomitant \nserum dosage of CA 125, CA 19–9 and interleukin-6 to detect \nthe presence of endometriosis. Results from a series of repro-\nductive age women undergoing laparoscopic surgery for benign \ngynaecological conditions. Hum Reprod 19:1871–1876. https://  \ndoi. org/ 10. 1093/ humrep/ deh312\nSuen J-L, Chang Y, Chiu P-R et al (2014) Serum level of IL-10 is \nincreased in patients with endometriosis, and IL-10 promotes the \ngrowth of lesions in a murine model. Am J Pathol 184:464–471. \nhttps:// doi. org/ 10. 1016/j. ajpath. 2013. 10. 023\nSwift B, Taneri B, Becker CM et al (2024) Prevalence, diagnostic delay \nand economic burden of endometriosis and its impact on quality \nof life: results from an Eastern Mediterranean population. Eur J \nPub Health 34:244–252. https:// doi. org/ 10. 1093/ eurpub/ ckad2 16\nSzabo E, Timmers I, Borsook D et al (2022) Altered anterior insula \nfunctional connectivity in adolescent and young women with \nendometriosis-associated pain: pilot resting-state fMRI study. \nEur J Paediatr Neurol 41:80–90. https:// doi. org/ 10. 1016/j. ejpn. \n2022. 10. 004\nSzubert M, Suzin J, Wierzbowski T, Kowalczyk-Amico K (2012) \nCA-125 concentration in serum and peritoneal fluid in patients \nwith endometriosis - preliminary results. Arch Med Sci 8:504–\n508. https:// doi. org/ 10. 5114/ aoms. 2012. 29407\nTaketani Y, Kuo TM, Mizuno M (1992) Comparison of cytokine lev -\nels and embryo toxicity in peritoneal fluid in infertile women \nwith untreated or treated endometriosis. Am J Obstet Gynecol \n167:265–270. https:// doi. org/ 10. 1016/ s0002- 9378(11) 91672-x\nTarokh M, GhaffariNovin M, Poordast T et al (2019) Serum and Peri-\ntoneal Fluid Cytokine Profiles in Infertile Women with Endome-\ntriosis. Iranian J Immunol 16:151–162. https:// doi. org/ 10. 22034/ \nIJI. 2019. 80258\nTu Q, Zhao R, Lu N (2024) Evaluation of the diagnostic utility of \nimmune microenvironment-related biomarkers in endo-\nmetriosis using multidimensional transcriptomic data. J \nAssist Reprod Genet 41:3213–3223. https:// doi. org/ 10. 1007/  \ns10815- 024- 03261-z\nVanhie A, Caron E, Vermeersch E et al (2024) Circulating microR -\nNAs as non-invasive biomarkers in endometriosis diagnosis—a \nsystematic review. Biomedicines 12:888. https:// doi. org/ 10. 3390/ \nbiome dicin es120 40888\nVeillat V, Carli C, Metz CN et al (2010) Macrophage migration inhibi-\ntory factor elicits an angiogenic phenotype in human ectopic \nendometrial cells and triggers the production of major angio-\ngenic factors via CD44, CD74, and MAPK signaling pathways. \nJ Clin Endocrinol Metab 95:E403-412. https:// doi. org/ 10. 1210/ \njc. 2010- 0417\nWaelkens DE, Vanhie A et al (2020) The use of antibody arrays in the \ndiscovery of new plasma biomarkers for endometriosis. Reprod \nSci 27:751–762. https:// doi. org/ 10. 1007/ s43032- 019- 00081-w\nWang W-T, Zhao Y-N, Han B-W et al (2013) Circulating micrornas \nidentified in a genome-wide serum microrna expression analysis \nas noninvasive biomarkers for endometriosis. J Clin Endocrinol \nMetab 98:281–289. https:// doi. org/ 10. 1210/ jc. 2012- 2415\nWang X-M, Ma Z-Y, Song N (2018) Inflammatory cytokines IL-6, \nIL-10, IL-13, TNF-α and peritoneal fluid flora were associated \nwith infertility in patients with endometriosis. European Rev \nMed Pharmacol Sci 22:2513–2518. https:// doi. org/ 10. 26355/ \neurrev_ 201805_ 14899\nWang H, Butler D, Zhang Y, Avery J, Knox S, Ma C, Hull L, Carneiro \nG (2024) Human-AI collaborative multi-modal multi-rater learn-\ning for endometriosis diagnosis. Phys Med Biol 70(1). https://  \ndoi. org/ 10. 1088/ 1361- 6560/ ad997e\nWeisheng B, Nezhat CH, Huang GF et al (2019) Discovering endome-\ntriosis biomarkers with multiplex cytokine arrays. Clin Proteom-\nics 16:28. https:// doi. org/ 10. 1186/ s12014- 019- 9248-y\nWerdel R, Mabie A, Evans TL et al (2024) Serum levels of interleukins \nin endometriosis patients: a systematic review and meta-analysis. \nJ Minim Invasive Gynecol 31:387-396.e11. https:// doi. org/ 10. \n1016/j. jmig. 2024. 02. 011\nWu M-H, Sun HS, Lin C-C et al (2002) Distinct mechanisms regulate \ncyclooxygenase-1 and -2 in peritoneal macrophages of women \nwith and without endometriosis. Mol Hum Reprod 8:1103–1110. \nhttps:// doi. org/ 10. 1093/ molehr/ 8. 12. 1103\nYang Y-T, Jiang X-Y, Xu H-L et al (2023) Autoimmune disease-related \nhub genes are potential biomarkers and associated with immune \nmicroenvironment in endometriosis. Int J Gen Med 16:2897–\n2921. https:// doi. org/ 10. 2147/ IJGM. S4174 30\nYun BH, Chon SJ, Choi YS et al (2016) Pathophysiology of endome-\ntriosis: role of high mobility group box-1 and toll-like recep -\ntor 4 developing inflammation in endometrium. PLoS ONE \n11:e0148165. https:// doi. org/ 10. 1371/ journ al. pone. 01481 65\nZamani P, Hemati Z, Kelishadi R et al (2023) Association between ano-\ngenital distance as a noninvasive index in the diagnosis and prog-\nnosis of reproductive disorder: A systematic review. Int J Reprod \nBiomed 21:599–618. https:// doi. org/ 10. 18502/ ijrm. v21i8. 14016\nZapardiel I, Gorostidi M, Ravaggi A et al (2016) Utility serum marker \nHE4 for the differential diagnosis between endometriosis and \nadnexal malignancy. Int J Gynecol Cancer 26:52–55. https:// doi. \norg/ 10. 1097/ IGC. 00000 00000 000579\nZhang W, Li K, Jian A et al (2025) Prospects for potential therapy \ntargeting immune-associated factors in endometriosis (Review). \nMol Med Rep 31:57. https:// doi. org/ 10. 3892/ mmr. 2024. 13422\nZhao B, Wang Z, Liu D, Zhang S (2023) Genetically predicted serum \ntestosterone and risk of gynecological disorders: a Mendelian \nrandomization study. Front Endocrinol (Lausanne) 14:1161356. \nhttps:// doi. org/ 10. 3389/ fendo. 2023. 11613 56\nZhao N, Hao T, Zhang F et al (2024) Application of machine learn-\ning techniques in the diagnosis of endometriosis. BMC Womens \nHealth 24:491. https:// doi. org/ 10. 1186/ s12905- 024- 03334-2\nPublisher's Note Springer Nature remains neutral with regard to \njurisdictional claims in published maps and institutional affiliations.","source_license":"CC-BY-4.0","license_restricted":false}