{"paper_id":"3f58807a-8955-4590-b780-59154712f6ce","body_text":"Endometriosis is a benign inflammatory gynecological \ndisease that manifests in women of reproductive age and \nis defined as the presence of stromal and viable endometrial \nglands outside the uterine cavity. It has an estimated \nprevalence of up to 10% in the general population and is \nassociated with chronic pelvic pain, dysmenorrhea and infertility \n( 1 ,  2 ). The pathogenic mechanisms, however, still \nremain unclear and the aetiology of the disease is believed \nto be multifactorial. Various endocrinological and immunological \nfactors have been investigated and are considered to \nsignificantly contribute to its pathophysiology ( 3 ,  4 ).\nDuring the past few years, several novel serum biomarkers \nhave been proposed for the early diagnosis of \nendometriosis. The most consistently studied molecule \nis cancer antigen 125 (CA-125), a glycoprotein that has \nbeen established as a biomarker for the follow-up of patients \nwith epithelial ovarian cancer. Similarly, in endometriosis, \nCA-125 can only be used as a prognostic rather \nthan a diagnostic marker as it is accompanied by a significant \namount of false negative results ( 5 ).\nRecently, urocortin (UCN) has been extensively investigated \nin the field of endometriosis. UCN is a neuropeptide \nthat belongs to the corticotrophin-releasing hormone \n(CRH) family and is expressed by eutopic and ectopic \nhuman endometria and is thought to play a role during \ndecidualization ( 6 ,  7 ). Because of its paracrine and immunomodulatory \nnature, UCN is thought to contribute to \nthe pathogenesis of endometriosis. To date, three different \nisoforms have been described (UCN1, UCN2 and UCN3), \nall of which exert their biological action by activating \ncorticotropin-releasing hormone (CRH) receptors 1 and \n2. Their main difference relies in the fact that UCN1 binds \nto both CRH1 and CRH2, whereas UCN2 and UCN3 bind \nselectively to CRH2 ( 8 ).\nTo date, it remains unclear whether UCN may be used \nas a screening and/or prognostic biomarker of endometriosis. \nThe objective of this systematic review is to accumulate \ncurrent evidence related to the expression of UCN \nin tissue and blood samples of patients suffering from \nendometriosis and provide directions for future research.\n\nThe present systematic review was designed according\nto the Preferred Reporting Items for Systematic Reviews \nand Meta-Analyses (PRISMA) guidelines ( 9 ). Eligibility \ncriteria were assessed by the authors. Briefly, date and \nlanguage restrictions were avoided during the literature \nsearch. All observational studies (both prospective and \nretrospective) that presented data relevant to the expression \nof UCN in tissue and blood samples of patients with \nendometriosis were included in this systematic review. \nStudies that defined their controls as either women with \nno pathology or women with other benign pathology were \nevaluated and included. Review articles, animal studies \nand case reports were excluded from the present review. \nThe selection process took place in three consecutive \nstages. Firstly, the titles and abstracts of all electronic \narticles were screened to assess their eligibility. Subsequently, \nthe articles that met or were presumed to meet \nthe criteria were retrieved as full texts. In the final stage \nreferences of articles that were retrieved in full text were \nevaluated to identify studies that might have been overlooked \nduring the electronic search. Any discrepancies in \nthe methodology, retrieval of articles and statistical analysis \nwere resolved through the consensus of all authors.\nLiterature search was primarily conducted using the \nMedline (1966-2018), Scopus (2004-2018), EMBASE \n(1947-2018) and Clinicaltrials.gov (2008-2018) databases, \nalong with the reference lists of electronically retrieved \nfull-text papers. Additional sources were identified through \nthe Google Scholar (2004-2018) database. Our last search \nwas on the 04 February 2018. The search strategy included \nthe words “endometriosis and urocortin” and is schematically \npresented in the PRISMA flow diagram ( Fig .1 ).\nSearch plot diagram.\n\nOverall, 8 studies were included in the present systematic \nreview and outcomes from a total of 567 women were \nassessed ( 10 - 18 ). The methodological characteristics of \nincluded studies are presented in Table 1. One study was \nexcluded from the present systematic review as it presented \npreliminary data that were solely based on immunohistochemistry \n( 13 ).\nKempuraj et al. ( 13 ) investigated in this pilot study the \nexpression of  UCN  in biopsies from 10 patients with endometriosis \nand 3 patients that did not have endometriosis. \nAlthough they did not perform statistical analysis on the \nobserved differences, they found that endometriotic lesions \nhad increased  UCN  expression compared with healthy peritoneum \nand normal endometrium. Two studies evaluated \n UCN  transcript expression ( 10 ,  18 ) and one study assessed \n UCN1  and  UCN2  transcript expression in eutopic endometrium \nand in endometriotic lesions. While  UCN2  expression \ndid not seem to differ between eutopic endometrium \nand endometriomas ( 16 ), the expression levels of  UCN  and \n UCN3  in the endometriotic foci (ectopic lesions) were significantly \nhigher compared with those in eutopic endometrium \nof the same women ( 10 ,  16 ,  18 ). Vergetaki et al. ( 18 ) \nshowed an almost 3.4-fold increase in the expression levels \nof  UCN  transcripts when ectopic and eutopic endometrium \nsamples were compared (0.9566 ± 0.136 a.u vs. 0.2826 ± \n0.075 a.u respectively).\nThe extent, depth of invasion and location of endometriotic \nlesions were shown to be associated with  UCN  transcript \nlevels. Specifically, Carrarelli et al. ( 10 ) observed \nthat deep infiltrating endometriosis (DIE) is associated \nwith higher levels of  UCN  than ovarian endometriomas \n(OMA) ( 10 ).\nNovembri et al. ( 15 ) studied  UCN, UCN2  and  UCN3  \ngene expression in eutopic endometrium of healthy \nwomen and women with endometriosis during the menstrual \ncycle. In women suffering from endometriosis, the \nexpression of  UCN, UCN2  and  UCN3  transcripts was \nthe same in the secretory and the proliferative phases, \nwhereas in healthy women  UCN  levels differed between \nthe two phases. Specifically,  UCN2  expression had peak \nvalues during the early proliferative phase, while  UCN3  \nexpression was at its maximum in the secretory phase ( 15 , \n 16 ). Both  UCN2  and  UCN3  expression levels were significantly \nlower in women with endometriosis when compared \nwith healthy women ( 16 ).\nDecidualization is a process of endometrial remodeling \nthat occurs in the secretory phase of the menstrual \ncycle and is essential for early pregnancy. Current \nknowledge suggests that this process is initiated \nby progesterone and is mediated by various molecules \nsuch as UCN. Novembri et al. ( 15 ) showed that women \nwith endometriosis have decreased levels of CRH and \nUCN, and suggested that this could negatively affect \ndecidualization.\nMethodological characteristics of included studies\nUCN; Urocortin, OMA; Ovarian endometrioma, DIE; Deeply infiltrating endometriosis,  * ; Patients with both OMA and DIE lesions were excluded, IHC; Immunohistochemistry, \nRT-PCR; Real-time polymerase chain reaction, and ELISA; Enzyme-linked immunosorbent assay.\nTwo studies evaluated preoperative plasma levels of \nUCN as a diagnostic factor that would help differentiate \npatients with endometriosis from patients with non-endometriotic, \nbenign ovarian cysts ( 12 ,  17 ). Specifically, Florio \net al. ( 12 ) reported that plasma UCN levels were two-\nfold higher in women with endometriomas (median 49 \npg/mL, interquartile range 41-63 pg/mL) compared with \ncontrols (19 pg/mL, P<0.001). The receiver operating \ncharacteristic (ROC) analysis showed that UCN detected \n88% of cases that had endometriomas with a specificity \nof 90% and an area under the curve (AUC) equal to 0.961 \n± 0.021 (cut-off value 33 pg/ml). Positive and negative \nlikelihood ratios for UCN were 8.8 and 0.14 respectively. \nOn the contrary, Tokmak et al. ( 17 ) reported no difference \nin the expression of UCN between patients with endometriomas \nand the control group (4.8 ± 1.00 ng/ml vs. 4.5 \n± 1.03 ng/ml, P=0.21). When the cut-off point was set at \n4.16 ng/ml, the sensitivity of the UCN protein in detecting \nendometriosis was 76.2%, the specificity was 45.7% and \nthe positive predictive value was 56.1%.\nChmaj-Wierzchowska et al. ( 11 ) compared UCN levels \nbetween patients with endometriomas and patients \nwith mature teratomas. The expression of UCN was not \nsignificantly different between the two groups (252.37 ± \n348.77 pg/ml vs. 256.03 ± 353.92 pg/ml, P=0.0727).\nMaia et al. ( 14 ) studied plasma levels of UCN1 as a \ndiagnostic biomarker of endometriosis among symptomatic \npatients. Compared with no-lesion patients (median \n34 pg/ml, interquartile range 22-43 pg/ml), patients \nwith endometriosis showed elevated UCN1 plasma levels \n(median 59 pg/ml, interquartile range 48-107 pg/ml). The \nROC analysis identified plasma UCN1 concentration of \n46 pg/mL as the best cut-off point to differentiate women \nwith endometriosis from those with no lesions, with 76% \nsensitivity, 88% specificity and an AUC equal to 0.827. \nHowever, an optimal cut-off that would distinguish endometriosis \nfrom other benign pathology (including benign \novarian cysts, ovarian teratoma, hydrosalpinx, salpingitis, \nectopic pregnancy, uterine leiomyoma and ovarian cancer) \nwas not identified.\n\nCurrent evidence suggests that UCN may be a promising \nfactor for the identification and follow-up of patients \nthat suffer from endometriosis. However, the methodological \nheterogeneity of these studies in terms of the reported \nmeasures precludes firm conclusions. The expression \nof UCN has been investigated post-transcription both \nat the transcription and protein levels. Three studies suggested \nthat the expression of  UCN  transcripts is significantly \nhigher in endometriotic lesions compared with eutopic \nendometrium of endometriotic women ( 10 ,  16 ,  18 ). \nIts correlation with the severity of the disease also implies \nthat it might become a useful tool for the classification of \nendometriosis ( 10 ). On the other hand, data related to the \nplasma levels of the protein were conflicting. Specifically, \ntwo studies reported significant differences between patients \nwith endometriosis and controls ( 12 ,  14 ), whereas \nanother two reported that UCN levels did not differ between \npatients with endometriomas and patients with \nother benign cysts ( 11 ,  17 ). Maia et al. ( 14 ) suggested that \nthe diagnostic accuracy of the expression at the protein \nlevel is promising, however, further evidence is needed to \nconfirm these findings.\nDespite the fact that our study is based on a meticulous \nreview of current literature, certain limitations preclude \ndefinitive conclusions. Firstly, the wide variation in outcome \nreporting and outcome reporting measures of UCN \nexpression in endometriosis precludes meta-analysis of \ncurrent data. Furthermore, the correlation between the \nexpression of UCN in endometriotic lesions and peripheral \nblood remains to be investigated. Its actual value as a \nminimally invasive diagnostic method therefore, remains \nto be elucidated. Future studies should also clarify whether \nUCN levels are elevated not only in endometriomas \nbut also in non-ovarian endometriosis and in early stage \ndisease. The presence of a biomarker with high sensitivity \nand specificity in these cases is particularly important for \nthe differential diagnosis since, to date, minimally invasive \nmethods are non-existent ( 19 ). Moreover, to standardise \nthe measurement of UCN levels in daily clinical \npractice, further investigation is necessary to determine \nwhether the peptide concentration is affected by factors \nincluding menstrual cycle, obesity, exercise, stress, diabetes \nmellitus and other chronic diseases ( 20 ). Accordingly, \nresearch should also focus on potential confounders \nthat may affect the plasma levels of this protein, including \ndiseases that may trigger chronic inflammation. To date, \nthere is insufficient evidence to suggest that UCN acts as \na mediator of inflammation, as its expression might have \nbeen an effect rather than a cause of inflammation ( 21 ).\nHowever, this interesting point deserves further investigation. \nMoreover, previous studies have also shown that \nUCN is produced by the liver and kidney of healthy animals \n( 22 ) but evidence in humans is still lacking. Therefore, \nthe evaluation of UCN in patients with hepatic and/\nor liver dysfunction needs to be elucidated to help exclude \nthese diseases as confounders. It would also be prudent \nto perform multivariate analyses to determine the impact \nof the various factors that were mentioned in this section \npotentially affecting UCN levels. Finally, cut-off values \nshould be introduced to investigate the predictive efficacy \nof UCN. These cut-offs should be based on previous proposed \nvalues that were mentioned in this systematic review \nto evaluate consistency of results. Optimal cut-offs \nshould also be reported to help future research in this field.\n\nCurrent evidence suggests that UCN may be a promising \nfactor for diagnosis and/or prognosis of patients that \nsuffer from endometriosis. However, available data are \nscarce and the findings of currently published studies remain \nto be validated. Specifically, future studies should \nexamine whether plasma and tissue levels of UCN correlate \nin patients. Moreover, evaluation of the predictive \naccuracy of UCN with the use of pre-specified cut-off values, \nmentioned in the present systematic review, is needed \nto assess the reproducibility of previous findings in the \nfield.","source_license":"CC0","license_restricted":false}