{"paper_id":"d6b698dd-f603-457b-968d-1e902a607be5","body_text":"Urinary IL-1β and plasma IFN-γ levels reveal \ncomplementary immune mechanisms in \nendometriosis\nEliana Dimova-Mihaylova1,2, Ivanka Dimova3 , Petia Andreeva4 , Dragomira Nikolova3\n1 Faculty of Medicine, Department of Biology, Medical Genetics, and Microbiology, Sofia University “St. Kliment Ohridski“, Sofia, Bulgaria\n2 University Hospital “Lozenetz”, Sofia, Bulgaria\n3 Faculty of Medicine, Department of Medical Genetics, Medical University–Sofia, Sofia, Bulgaria\n4 Medical Complex “Dr. Shterev”, Sofia, Bulgaria\nCorresponding author: Dragomira Nikolova (dmb@abv.bg)\nReceived 24 February 2026   ♦   Accepted 23 April 2026   ♦   Published 13 May 2026\nCitation: Dimova-Mihaylova E, Dimova I, Andreeva P , Nikolova D (2026) Urinary IL-1β and plasma IFN-γ levels reveal comple -\nmentary immune mechanisms in endometriosis. Pharmacia 73: e189520. https://doi.org/10.3897/pharmacia.73.e189520\nAbstract\nObjectives: Endometriosis is a multifactorial inflammatory disorder characterized by the coexistence of localized inflammatory ac-\ntivation and systemic immune dysregulation. Single biomarkers are unlikely to capture this biological complexity, underscoring the \nneed for integrated immune profiling across different biological compartments. This pilot study evaluated interleukin-1β (IL-1β) and \ninterferon-γ (IFN-γ) in paired plasma and urine samples from women with clinically confirmed endometriosis and healthy controls \nto explore compartment-specific cytokine patterns.\nMethods: IL-1β and IFN-γ were quantified in plasma and urine samples from patients with endometriosis and healthy control \nwomen using the automated microfluidic ELISA system (Ella, ProteinSimple/Bio-Techne, USA). Marker detectability and expression \npatterns were evaluated individually. The expression levels were compared using the Mann–Whitney U test.\nResults: IL-1β was undetectable in plasma samples in both groups but was detected in urine samples from patients with endometri-\nosis but not in controls (p = 0.0303). In contrast, IFN-γ was detectable in plasma but not in urine. Plasma IFN-γ levels were lower in \npatients compared with controls; however, this difference was not statistically significant (p > 0.99) and should be interpreted only \nas a descriptive trend.\nConclusion: The findings of this study suggest compartment-specific immune differences in endometriosis, with urinary IL-1β \nreflecting localized inflammatory activity and plasma IFN-γ showing a trend toward reduced systemic T helper 1 (Th1)-associated \nimmunity. Larger studies are required to validate these preliminary observations. This complementary biomarker profile underscores \nthe importance of integrated interpretation across biological systems and supports a minimally invasive immune signature for im -\nproved characterization of endometriosis.\nKeywords\nBiomarkers, endometriosis, noninvasive\nCopyright Dimova-Mihaylova E et al. This is an open access article distributed under the terms of the Creative Commons Attri-\nbution License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original \nauthor and source are credited.\nPharmacia 73: e189520\nDOI 10.3897/pharmacia.73.e189520\nResearch Article\n\nDimova-Mihaylova E, et al.: Potential cytokine biomarkers of endometriosis\n2\nIntroduction\nEndometriosis is a chronic, estrogen-dependent in -\nflammatory disease that affects an estimated 10–15% of \nwomen of reproductive age worldwide and represents a \nmajor cause of pelvic pain, dysmenorrhea, and infertility \n(Garmendia et al. 2025). Despite its significant impact on \nquality of life, diagnosis remains challenging and is often \ndelayed for several years, largely due to the absence of re-\nliable noninvasive biomarkers (Agic et al. 2006; Garmen-\ndia et al. 2025). This diagnostic gap reflects the biological \ncomplexity of the disease, which arises from the interplay \nbetween hormonal regulation and immune dysfunction \nrather than a single dominant pathogenic pathway.\nOne of the prevailing theories for the occurrence of \nendometriosis is the underlying process of immune dys -\nregulation, which contributes both to lesion establishment \nand persistence. Alterations have been described across \nmultiple components of the immune system, including \nenhanced macrophage activation, impaired natural killer \n(NK) cell cytotoxicity, imbalances in T helper cell subsets, \nand aberrant B-cell responses (Abramiuk et al. 2022; Vall-\nvé-Juanico et al. 2022). Together, these changes promote \nan inflammatory microenvironment that favors ectopic \ntissue survival while limiting effective immune clearance. \nWithin this context, cytokines play a pivotal role as me -\ndiators of immune communication, tissue remodeling, \nangiogenesis, and pain signaling (Bergqvist et al. 2001; \nUçeyler et al. 2007; Turner et al. 2014).\nNumerous studies have reported elevated concentra -\ntions of pro-inflammatory cytokines—including inter -\nleukin-1 beta (IL-1β), interleukin-6 (IL-6), and tumor \nnecrosis factor-α (TNF-α)—in endometriotic lesions and \nperitoneal fluid, with less consistent findings in peripheral \nblood (Bergqvist et al. 2001; Krygere et al. 2024). However, \ntranslation into clinically useful biomarkers has been lim-\nited due to variability across menstrual cycle phase, lesion \nlocalization, disease stage, and biological compartment. \nCytokine measurements are also influenced by systemic \ninflammatory conditions and lack disease specificity, lim-\niting the utility of single-marker approaches (Agic et al. \n2006). As a result, single cytokine measurements often fail \nto reflect the underlying complexity of the disease.\nMost immunological studies in endometriosis focus on \nserum and peritoneal fluid, where numerous cytokines, \nsuch as IL-6, IL-8, TNF-α, and MCP-1, are elevated com-\npared with controls and correlate with disease activity \nor severity (Wu and Ho 2003). There are very few direct \nstudies measuring classical cytokines, such as IL-6, IL-8, \nMCP-1, and TNF-α, in urine from patients with endome-\ntriosis as diagnostic biomarkers (Liu et al. 2015). Some \nstudies have measured solitary urinary proteins related to \ninflammation, such as cytokeratin-19 fragment (CYFRA \n21-1) (Saputra et al. 2025), urinary vitamin D-binding \nprotein (Cho et al. 2012), urocortin, and other non-cy -\ntokine markers (Toma et al. 2025). Urine has long been \nshown to be a reliable source of diagnostic material; how-\never, classical cytokines are often present at very low con-\ncentrations or below detection limits. Evidence about the \nlevels of classical cytokines in urine is so far limited and \ninconclusive.\nAmong the immune mediators of endometriosis, IL-1β \nand IFN-γ represent distinct and complementary aspects \nof immune function. IL-1β is a key driver of innate inflam-\nmatory responses and is produced primarily by activated \nmacrophages. It has been repeatedly shown to be upregu-\nlated in endometriosis tissue and peritoneal fluid, where it \npromotes leukocyte recruitment, angiogenesis, and local \ninflammation (Sikora et al. 2016). In contrast, IFN-γ is a \nhallmark cytokine of T helper 1 (Th1)-mediated immuni-\nty and plays a critical role in cellular immune surveillance \nand macrophage activation (Castro et al. 2018).\nIn this study, IL-1β and IFN-γ levels were measured in \npaired urine and plasma samples obtained from individuals \nwith endometriosis and from controls without the disease. \nBy evaluating those two protein markers, this study aims \nto identify a complementary immune signature that reflects \nboth local and systemic aspects of endometriosis. This ap-\nproach explores the potential of cytokine patterns as a mini-\nmally invasive strategy for improving disease identification.\nMaterials and methods\nStudy participants\nA total of 16 women were included in this study: 10 pa -\ntients with endometriosis (mean age 34 years) and six \nhealthy control women of the same ethnicity (mean age 22 \nyears). All participants provided written informed consent \nfor participation in the study, and the study protocol was \napproved by the Ethical Committee of the Scientific Coun-\ncil of the Medical University—Sofia, Bulgaria (Approval No. \n2826/24.06.2025, Protocol No.13/05.06.2025). All women \nwere diagnosed with confirmed, non-surgically treated en-\ndometriosis and were selected at the University Hospital \n“Lozenets” or Medical Complex “D-r Shterev” (Table 1).\nSample collection and storage\nPeripheral venous blood was collected in EDTA tubes and \ncentrifuged at 2,000× g for 10 mins to separate the plasma. \nPlasma samples were aliquoted and stored at –80 °C until \nanalysis. Morning midstream urine samples were collect-\ned, aliquoted, and similarly stored at –80 °C.\nCytokine quantification using the Ella® \nsystem\nPlasma and urine concentrations of IL-1β and IFN-γ (pg/\nmL) were measured using the Ella automated immunoas-\nsay system (ProteinSimple/Bio-Techne, MN, USA). The Ella \nsystem is a fully automated, microfluidic-based platform \nthat performs multiplexed sandwich ELISA assays in nano-\nliter-scale volumes. The system reports mean RFU (relative \nfluorescence units), RFU %CV , mean concentration, and \nconcentration %CV , enabling direct quantification of cyto-\nkine concentrations with built-in quality control. Each sam-\nple was assayed at a 1:2 dilution for both plasma and urine.\n\nPharmacia 73: e189520 3\nData presentation and statistical anal -\nysis\nIL-1β and IFN-γ levels were quantified in plasma and \nurine samples from 10 women with endometriosis and six \nhealthy controls and are presented as mean ± standard de-\nviation (SD). Due to the limited sample size and non-nor-\nmal distribution of the data, between-group comparisons \nof protein concentrations (pg/mL) were performed using \nthe non-parametric Mann–Whitney U test. Statistical \nanalyses were conducted using GraphPad Prism version \n10.6.1 (GraphPad Software, Boston, MA, USA). A p-value \n< 0.05 was considered statistically significant.\nResults below the lower limit of quantification (LLOQ) \nwere replaced with LLOQ/2. Results with a coefficient of \nvariation > 25% were excluded.\nResults\nIL-1β and IFN-γ in plasma and urine\nIL-1β was undetectable in plasma samples from both pa -\ntients and controls.\nUrinary IL-1β levels differed between groups (p = 0.0303), \nwith higher levels observed in women with endometriosis \n(mean = 4.99 pg/mL) compared with controls (mean = 0.75 \npg/mL). The distribution of urinary IL-1β concentrations \nmeasured using the Ella® system is shown in Fig. 1.\nIFN-γ was not detectable in urine samples from ei -\nther group.\nIn plasma, IFN-γ was detectable in both groups. Mean \nlevels were lower in patients with endometriosis compared \nwith controls (0.70 pg/mL vs 1.54 pg/mL). However, this \ndifference was not statistically significant (p > 0.99) and is \ntherefore reported only as a descriptive observation.\nThe plasma IFN-γ concentrations are shown in Fig. 2.\nDiscussion\nIn this study, the protein expression of IL-1β and IFN-γ \nwas investigated in paired plasma and urine samples from \nwomen with endometriosis and healthy controls. The find-\nings suggest differential cytokine detectability depending \non the studied biological fluid, as well as on disease status.\nIL-1β appeared undetectable in plasma samples from \nboth groups, supporting the concept that endometriosis \nis primarily characterized by localized rather than overt \nsystemic inflammation (Giudice and Kao 2004; Bulun et \nal. 2009). In contrast, IL-1β was vaguely detectable in urine \nsamples from women with endometriosis, whereas it re -\nmained largely undetectable in controls. Despite the small \nTable 1. Clinical and demographic characteristics of the study \ngroups.\nNumber Sample code Age Diagnosis\nPatients\n1 P1-PR 37 endometriosis\n2 P2-PR 28 endometriosis\n3 P3-PR 27 endometriosis\n4 P4-PR 32 endometriosis\n5 P9-PR 43 endometriosis\n6 PR-P10 37 endometriosis\n7 PR-P15 34 endometriosis\n8 PR-P16 34 endometriosis\n9 PR-P17 36 endometriosis\n10 PR-P20 32 endometriosis\nControls\n1 PR-C2 27 control group\n2  PR-C10 19 control group\n3 PR-C11 21 control group\n4 PR-C12 20 control group\n5 PR-C13 21 control group\n6 PR-C17 22 control group\nFigure 1.  The distribution of urinary IL-1β concentrations \nmeasured by the automated ELISA system Ella® (ProteinSimple/\nBio-Techne, USA) in pg/mL in patients with endometriosis and \ncontrol individuals. * indicates statistical significance between \nthe analyzed groups (p = 0.0303).\nFigure 2.  The distribution of plasma IFN-γ concentrations \nmeasured by the automated ELISA system Ella® (Bio-Techne, \nBoston, USA) in pg/mL in patients with endometriosis and con-\ntrol individuals. “ns” indicates a lack of statistical significance \nbetween the analyzed groups (p > 0.99).\n\n\nDimova-Mihaylova E, et al.: Potential cytokine biomarkers of endometriosis\n4\nnumber of samples, this observation demonstrated a clear \ntrend toward higher urinary IL-1β levels in patients with \nendometriosis. This is consistent with the participation of \nIL-1β in inflammasome activation, angiogenesis, and in -\nflammatory amplification within the pelvic microenviron-\nment (Lebovic et al. 2001; Dinarello et al. 2011). IL-1β has \nbeen consistently reported as elevated in the peritoneal \nfluid of women with endometriosis, where it contributes \nto immune cell recruitment (Lebovic et al. 2001; Sikora et \nal. 2018). The detection of IL-1β in urine may be caused by \nlocalized pelvic inflammation, potentially offering a non-\ninvasive way to monitor the inflammatory activity associ-\nated with the disease.\nIn contrast, IFN-γ was undetectable in urine in both \ngroups, suggesting limited urinary excretion or absent lo-\ncal production in this compartment. Plasma IFN-γ levels \nwere lower in patients with endometriosis compared with \ncontrols; however, this difference was not statistically sig-\nnificant and should be interpreted cautiously as a possible \nrather than a confirmed biological effect. IFN-γ is predom-\ninantly produced by Th1 CD4+ T cells, cytotoxic CD8+ T \ncells, and natural killer (NK) cells and directs macrophage \nactivation and enhanced antigen presentation, which are \nessential for the effective clearance of aberrant cells (Cas-\ntro et al. 2018; Alspach et al. 2019). Relatively lower sys -\ntemic IFN-γ levels in patients, if confirmed, could reflect a \nshift away from Th1 immunity, which is mainly implicated \nin endometriosis pathogenesis, particularly through NK \ncell dysfunction and reduced cytotoxicity against ectopic \nendometrial tissue (Reis et al. 2022; Chen et al. 2023).\nIFN-γ cannot currently be considered a candidate bio-\nmarker in this dataset, and the trend toward lower plasma \nIFN-γ in patients warrants further investigation, partic -\nularly using larger cohorts or in combination with func -\ntional assays to assess Th1 and NK cell activity. This may \nbe of interest for future hypothesis-driven studies examin-\ning systemic Th1 activity in endometriosis.\nLimitations of the study\nImportantly, several limitations must be acknowledged. \nThe study comprises a relatively small cohort (10 patients \nand six controls), which limits the statistical power and \nincreases the likelihood of type II error. The observed sta-\ntistical significance for urinary IL-1β should be interpret-\ned with caution. Even if the observed trend is biologically \nconsistent with the current understanding of localized in-\nflammatory activation in endometriosis, it should be vali-\ndated in a larger number of samples.\nSeveral clinically relevant variables that may influence \ncytokine levels were not controlled in this study, including \nmenstrual cycle phase, hormonal therapy status, BMI, dis-\nease stage, and lesion localization. These factors are known \nto significantly affect immune signaling and cytokine ex -\npression in endometriosis. Future studies should incorpo-\nrate stratification for these variables.\nStrengths of the study\nDespite these limitations, the present study has several \nnotable strengths. The parallel assessment of cytokines \nin two biological compartments (plasma and urine) pro -\nvides a compartment-specific perspective that is rarely \nexplored in endometriosis research. Most previous stud -\nies have focused on serum or peritoneal fluid; therefore, \nthe inclusion of urine expands the perspectives toward \nthe evaluation of a noninvasive sampling matrix. The \nfindings are biologically coherent with the current un -\nderstanding of endometriosis as a localized inflammato -\nry disorder. The selective detection of IL-1β in urine from \npatients aligns with established evidence of IL-1–medi -\nated pelvic inflammation and inflammasome activation \n(Lebovic et al. 2001; Dinarello et al. 2011).\nConclusion\nIn summary, this pilot study suggests a distinct compart -\nment-specific immune pattern in endometriosis, charac -\nterized by detectable urinary IL-1β and a nonsignificant \nreduction in plasma IFN-γ. This combination reflects the \ndual nature of the disease, involving localized pelvic in -\nflammation alongside altered systemic Th1-mediated im-\nmune responses. Larger, controlled studies are required \nto validate these observations before any possible clinical \nutility consideration and application.\nAdditional information\nConflict of interest\nThe authors have declared that no competing interests exist.\nEthical statements\nThe authors declared that no clinical trials were used in the pres-\nent study.\nAll participants provided written informed consent for par -\nticipation in the study, and the study was approved by the Eth -\nics Committee of the Medical University of Sofia (Approval No. \n2826/24.06.2025, Protocol No.13/05.06.2025)\nThe authors declared that no experiments on animals were \nperformed in the present study. \nThe authors declared that no commercially available immor-\ntalized human or animal cell lines were used in the present study.\nArtificial Intelligence (AI) use\nThe authors accept full responsibility for the content of the man-\nuscript, including the disclosure of any use of AI.\nNo AI tools were used in the preparation of this manuscript.\nFunding\nThis study is financed by project № D-156/04.06.2025 of the Coun-\ncil of Medical Sciences, Medical University–Sofia, Sofia, Bulgaria.\n\nPharmacia 73: e189520 5\nAuthor contributions\nE.Dimova-Mihaylova has collected part of the samples at \n“Lozenetz” hospital, has performed the experiments, has made \nthe statistical analysis; I.Dimova has participated in the statistical \nanalysis of the data and has suggested some interpretations; P .An-\ndreeva has collected the blood/urine samples at the Medical com-\nplex “D-r Shterev” and has selected the patients; D.Nikolova has \nparticipated in the experiments and analysis, suggested improve-\nments, interpretations and critically revised the manuscript.\nAuthor ORCIDs\nIvanka Dimova  https://orcid.org/0000-0003-4091-9604\nPetia Andreeva  https://orcid.org/0000-0002-6277-8550\nDragomira Nikolova  https://orcid.org/0000-0002-8929-8522\nData availability\nAll of the data that support the findings of this study are available \nin the main text.\nReferences\nAbramiuk M, Grywalska E, Malkowska P , Sierawska O, Hrynkiewicz R, \nNiedzwiedzka-Rystwej P (2022) The role of the immune system in \nthe development of endometriosis. Cells 11(13): 2028. https://doi.\norg/10.3390/cells11132028\nAgic A, Xu H, Finas D, Banz C, Diedrich K, Hornung D (2006) Is en -\ndometriosis associated with systemic subclinical inflammation? \nGynecologic and Obstetric Investigation 62(3): 139–147. https://doi.\norg/10.1159/000093121\nAlspach E, Lussier DM, Schreiber RD (2019) Interferon γ and its import-\nant roles in promoting and inhibiting spontaneous and therapeutic \ncancer immunity. Cold Spring Harbor Perspectives Biology 11(3). \nhttps://doi.org/10.1101/cshperspect.a028480\nBergqvist A, Bruse C, Carlberg M, Carlström K (2001) Interleukin 1beta, \ninterleukin-6, and tumor necrosis factor-alpha in endometriotic \ntissue and in endometrium. Fertility and Sterility 75(3): 489–495. \nhttps://doi.org/10.1016/s0015-0282(00)01752-0\nBulun SE (2009) Endometriosis. The New England Journal of Medicine \n360(3): 268–279. https://doi.org/10.1056/NEJMra0804690\nCastro F , Cardoso AP , Gonçalves RM, Serre K, Oliveira MJ (2018) \nInterferon-gamma at the crossroads of tumor immune surveil -\nlance or evasion. Frontiers in Immunology 9: 2028. https://doi.\norg/10.3389/fimmu.2018.00847\nChen S, Liu Y , Zhong Z, Wei C, Liu Y , Zhu X (2023) Peritoneal immune \nmicroenvironment of endometriosis: Role and therapeutic perspec -\ntives. Frontiers in Immunology 14: 2023. https://doi.org/10.3389/\nfimmu.2023.1134663\nCho S, Choi YS, Yim SY , Y ang HI, Jeon YE, Lee KE, Lee BS (2012) \nUrinary vitamin D-binding protein is elevated in patients with en -\ndometriosis. Human Reproduction 27(2): 515–522. https://doi.\norg/10.1093/humrep/der345\nDinarello CA (2011) Interleukin-1 in the pathogenesis and treatment \nof inflammatory diseases. Blood 117(14): 3720–3732. https://doi.\norg/10.1182/blood-2010-07-273417\nGarmendia JV , De Sanctis CV , Hajduch M, De Sanctis JB (2025) Endo-\nmetriosis: An immunologist’s perspective. Internation Journal of Mo-\nlecular Sciences 26(11): 5193. https://doi.org/10.3390/ijms26115193\nGiudice LC, Kao LC (2004) Endometriosis. The Lancet 364(9447): 1789–\n1799. https://doi.org/10.1016/S0140-6736(04)17403-5\nKrygere L, Jukna P , Jariene K, Drejeriene E (2024) Diagnostic poten -\ntial of cytokine biomarkers in endometriosis: Challenges and in -\nsights. Biomedicines 12(12): 2867. https://doi.org/10.3390/biomed-\nicines12122867\nLebovic DI, Mueller MD, Taylor RN (2001) Immunobiology of endo -\nmetriosis. Fertility and Sterility 75(1): 1–10. https://doi.org/10.1016/\ns0015-0282(00)01630-7\nLiu E, Nisenblat V , Farquhar C, Fraser I, Bossuyt PM, Johnson N, Hull \nML (2015) Urinary biomarkers for the non-invasive diagnosis of \nendometriosis. Cochrane Database of Systematic reviews 2015(12). \nhttps://doi.org/10.1002/14651858.cd012019\nReis JL, Rosa NN, Ângelo-Dias M, Martins C, Borrego LM, Lima J \n(2022) Natural killer cell receptors and endometriosis: A systematic \nreview. Internation Journal of Molecular Sciences 24(1): 331. https://\ndoi.org/10.3390/ijms24010331\nSaputra NPK, Samsulhadi S, Hendarto H, Sa’adi A, Pudjirahardjo WJ, Arsa-\nna IW , Y anuari R, Dwiningsih SR (2025) Strengthening of CYFRA 21–1 \nusing urine creatinine correction as potential endometriosis biomarker. \nF1000Research 13: 46. https://doi.org/10.12688/f1000research.135167.2\nSikora J, Mielczarek-Palacz A, Kondera-Anasz Z (2016) Association of \nthe precursor of interleukin-1β and peritoneal inflammation-role in \npathogenesis of endometriosis. Journal of Clinical Laboratory Anal-\nysis 30(6): 831–837. https://doi.org/10.1002/jcla.21944\nSikora J, Smycz-Kubańska M, Mielczarek-Palacz A, Bednarek I, Konde-\nra-Anasz Z (2018) The involvement of multifunctional TGF-β and \nrelated cytokines in pathogenesis of endometriosis. Immunology \nLetters 201: 31–37. https://doi.org/10.1016/j.imlet.2018.10.011\nToma B, Caruntu ID, Simionescu N, Onofriescu M, Socolov D, Ilea C, \nSocolov R (2025) Urinary urocortin as a potential non-invasive bio -\nmarker in endometriosis: Exploratory study with histone H4. Me -\ndicina 61(9): 1671. https://doi.org/10.3390/medicina61091671\nTurner MD, Nedjai B, Hurst T, Pennington DJ (2014) Cytokines and \nchemokines: At the crossroads of cell signalling and inflammatory dis-\nease. Biochimica et Biophysica Acta (BBA) – Molecular Cell Research \n1843(11): 2563–2582. https://doi.org/10.1016/j.bbamcr.2014.05.014\nUçeyler N, Rogausch JP , Toyka KV , Sommer C (2007) Differential expres-\nsion of cytokines in painful and painless neuropathies. Neurology \n69(1): 42–49. https://doi.org/10.1212/01.wnl.0000265062.92340.a5\nVallvé-Juanico J, George AF , Sen S, Thomas R, Shin MG, Kushnoor D, \nGiudice LC (2022) Deep immunophenotyping reveals endometriosis \nis marked by dysregulation of the mononuclear phagocytic system in \nendometrium and peripheral blood. BMC Med 20(1): 158. https://\ndoi.org/10.1186/s12916-022-02359-4\nWu MY , Ho HN (2003) The role of cytokines in endometriosis. Amer -\nican Journal of Reproductive Immunology 49(5): 285–296. https://\ndoi.org/10.1034/j.1600-0897.2003.01207.x","source_license":"CC0","license_restricted":false}