{"paper_id":"b0c535fc-8aea-46b0-8c28-0ce11d55f605","body_text":"Endometriosis is a gynecological disorder affecting the wellbeing of 5%-15% of women\nof reproductive age, with a prevalence of 5%-50% in infertile women and 32% in women\nwith chronic pelvic pain ( Bellelis  et\nal ., 2014 ). Studies showed that 30% to 50% of the women with\nendometriosis are infertile ( Zhang  et\nal ., 2014 ;  Zhu  et\nal ., 2014 ). Endometriosis is the presence of endometrial\ntissue (glandular epithelium and stroma) outside the normal location ( Abu Hashim, 2014 ). An estimated seven million\nwomen have endometriosis in the USA, and the disease ranks as one of the main causes\nfor gyneco logical hospitalization in industrialized nations. Gao  et\nal.  reviewed the direct medical and nonmedical burden associated with\nendometriosis. The authors suggested that endometriosis places a considerable burden\non patients and society ( Bellelis  et\nal ., 2014 ;  Gao  et\nal ., 2006 ).\nAccording to the definition of the American Society of Reproductive Medicine (ASRM),\nendometriosis can be categorized into four stages: stage I (minimal), stage II\n(mild), stage III (moderate), and stage IV (severe) ( Pacchiarotti  et al ., 2014 ). More advanced stages may be\ndeeply invasive and present as endometrioma ( Avcioğlu  et al ., 2014 ). Nearly a third (32%) of\nthe patients with endometriosis have moderate to severe disease, while 58% have\nminimal or mild endometriosis. The pathogenesis of mild/minimal endometriosis with\ninfertility is unclear ( Zhu  et al .,\n2014 ). Diagnostic laparoscopy, with or without biopsy for histological\ndiagnosis, is the most common procedure used to diagnose and remove mild to moderate\nendometriosis. This method is considered the gold standard among scoring systems\navailable for determining disease severity ( Marchino\n et al ., 2005 ).\nInvestigators have suggested that women with mild to moderate endometriosis have a\nhigher incidence of endocrine abnormalities, anovulation, and hyperprolactinemia.\nHowever, other well-organized prospective studies have found most of these factors\nto be either normal or lacking in clinical significance ( Gardner  et al ., 2017 ). Nevertheless, several\nclinical and experimental reports have suggested a relationship between\nendometriosis and its progression with hyperprolactinemia. There is controversy as\nto whether abnormal prolactin secretion is directly involved in infertility in\npatients with endometriosis ( Gardner  et\nal ., 2017 ;  Esmaeilzadeh\n et al ., 2015 ).\nThe real mechanisms of infertility associated with endometriosis in patients with\nhyperprolactinemia have not been entirely clarified. Regardless of the\ninterventional role of hyperprolactinemia in the endocrine pattern of infertility,\nit probably impairs luteinizing hormone (LH) pulsation and induces infertility\nthrough ovulation failure, luteinized unruptured follicle (LUF) syndrome or poor\nendometrial response to estrogen ( Wang  et\nal ., 2009 ). The studies that suggested a relationship\nbetween endometriosis and abnormal prolactin secretion are limited in number, and\ntheir results are controversial. In addition, far too little attention has been\ngiven to studies comparing prolactin levels in various stages of endometriosis.\nIn a previous study ( Esmaeilzadeh  et\nal ., 2015 ) we found a relationship between endometriosis and\nprolactin levels; in this study, we looked into whether hyperprolactinemia is a\nprobable prognostic biomarker to detect the severity of the endometriosis (minimal\nto severe) by analyzing further samples.\n\nThis is a retrospective cohort study. The data sets used herein were extracted from\nthe medical records of patients seen at the Infertility and Reproductive Health\nResearch Center at Babol University of Medical Science from January 2015 to\nSeptember 2016. The data collected included age, reasons and duration of\ninfertility, stages of endometriosis, serum prolactin (PRL) levels, and\nultrasound/laparoscopy findings.\nSerum PRL was measured with DiaSorin kits manufactured in Spain and with the aid of a\nLIAISON system using chemiluminescence technology (CLIA). PRL secretion was deemed\nnormal when baseline serum levels were 25ng/ml or lower at least two hours after\nwaking up in the morning ( Melmed  et\nal ., 2011 ). Patients categorized as having\nhyperprolactinemia had to have two prolactin level readings ≥25 ng/ml on the\nsecond or third day in two consecutive periods. Also, a patient was considered in\nthe normal prolactin group if the level of prolactin was normal at once. Since our\npatients had infertility and irregular menstrual periods, endocrine tests were run\nto exclude other potential ovarian endocrine defects that might have affected their\nstatus of infertility associated with endometriosis. The research project was\napproved by the Ethics Committee of the Babol University of Medical Science and\nwritten consent was obtained from all participants.\nThe group with endometriosis contained infertile patients with endometriosis\nconfirmed by laparoscopic examination. They were further segregated into two\nsubgroups, stage I/II endometriosis and stage III/IV endometriosis, and all\npatients with endometriosis were scored according to the World Endometriosis\nSociety consensus on the classification of endometriosis ( Johnson  et al ., 2017 ). The infertile\npatients in the group they were compared against underwent laparoscopic\nexamination and had no signs of endometriosis.\nExcluded patients were older than 40 years, had diseases such as thyroid\ndysfunction or renal disease or were taking drugs that caused\nhyperprolactinemia, or were in non-fasting conditions, exercised excessively,\nhad trauma, renal disease or inadequate data for analysis.\nStatistical analysis was performed on SPSS 19.0. The data were tested for\nnormality with the Kolmogorov-Smirnov test and were presented as mean values\n± (SD) or percentages when appropriate. The independent t-test was used\nto compare between baseline PRL levels of the two groups; the chi-squared test\nwas used to determine the relationships between categorical variables; and the\nMann-Whitney Test was used to compare the sample mean values coming from one\nsame group. Linear regression and logistic regression were used to determine the\nassociation between prolactin levels and stages of endometriosis. All tests were\ntwo-tailed and significant differences had a  p -value of less\nthan 0.05.\n\nOne hundred and twenty-three women were diagnosed with endometriosis. Nine were\nexcluded for different reasons (three were on pills for thyroid disorder; two were\nnot accessible; and three chose not to join the study). Of the 114 patients with\nendometriosis enrolled in the study, 37 (32.4%) had disease stages I/II (5 with\nstage I, 32 with stage II) and 77 (67.5%) had endometriosis stages III/IV (38 with\nstage III, 39 with stage IV). One hundred and one patients were included in the\ncontrol group. No one from the control group was excluded. There were no\nstatistically significant differences in age, level of education, body mass index or\nprimary infertility between the endometriosis and control groups ( Table 1 ).\nStudy population\nStudent’s  t -test\nχ 2\nData presented as mean values ± (SD).\nData presented as n (%)\nThe hormonal assay results of both groups are presented in  Table 2 . The mean PRL level was 17.88±12.81 ng/mL in the\ncontrol group ( Table 2 ); 23.42±34.05\nng/mL in the group with disease stages I and II; and 31.62±38.09 ng/mL in the\ngroup with disease stages III and IV. Serum prolactin levels were significantly\nhigher among infertile women with endometriosis than in infertile women without\nendometriosis ( p =0.003) ( Table\n2 ). Welch’s test revealed significant differences between the three\ngroups ( p =0.018). Tamhane’s multiple comparison test revealed a\nsignificant difference between controls and individuals with disease stages III/IV\n( p =0.009). The related  p -value indicates that\nthere is a 0.009 probability that chance produced in the relation of endometriosis\nstages III/IV and prolactin value, however the calculated effect size of the study\nwith 80% confidence showed -0.55 with CI (-0.74 - -0.35).\nHormonal assay results of women with and without endometriosis\nMean values ± SD\nThe differences between controls vs. subjects with disease stages I/II and stages\nI/II vs. stages III/IV were not significant ( p =0.71,\n p =0.58). The calculation of the effect size of stages I/II vs.\nstages III/IV showed that the strength of association was -0.23 with 80% CI (-0.42 -\n-0.03). The prognostic capability of prolactin in detecting endometriosis stages\nIII/ IV vs. stages I/II was analyzed by ROC curve analysis ( Figure 1 ). Prognostic capability was achieved with an AUC=0.65,\n95% CI (0.55, 0.76). Prolactin values with a cut off set at 20.08 ng/mL had a\nsensitivity of 0.61 and specificity of 0.60 to detect endometriosis stages III/IV\nvs. I/II. In addition, the prognostic capability of prolactin in segregating\nendometriosis cases from controls was identified by ROC curve analysis with an\nAUC=+0.67, 95% CI (0.60, 0.74). Prolactin levels of 17.5 ng/mL had a sensitivity of\n0.64 and specificity of 0.63 to differentiate endometriosis cases from controls.\nFigure 1 A. ROC Curve to assess the diagnostic capability of prolactin in\ndifferentiating between patients with and without endometriosis. B. ROC\nCurve to assess the diagnostic capability of prolactin in\ndifferentiating patients with endometriosis stages III/IV from patients\nwith endometriosis stages I/II\nA. ROC Curve to assess the diagnostic capability of prolactin in\ndifferentiating between patients with and without endometriosis. B. ROC\nCurve to assess the diagnostic capability of prolactin in\ndifferentiating patients with endometriosis stages III/IV from patients\nwith endometriosis stages I/II\n\nOur previous study revealed that infertile patients with endometriosis had\nhyperprolactinemia ( Esmaeilzadeh  et\nal ., 2015 ). In the present study, the most striking result\nto emerge from our data is that we found a good prognostic capability of prolactin\nin detecting patient with endometriosis stages III/IV vs. stages I/II. A clinical\nimplication of these findings is proposing hyperprolactinemia (Prolactin values\n≥20.08 ng/mL) as a probable prognostic biomarker for infertile women with\nendometriosis stages III/IV vs. I/II. We were unable to find a study that reported a\ncutoff value for prolactin in women with endometriosis stages III/IV vs. I/II.\nMoreover, the present study demonstrated that the serum prolactin levels of\ninfertile women with endometriosis stages III/IV were significantly higher than the\nlevels seen in infertile women without endometriosis. This finding is supported by\nprior studies ( Barbosa  et al .,\n2014 ;  Lima  et al .,\n2006 ;  Gregoriou  et al .,\n1999 ;  Cunha-Filho  et\nal ., 2001 ). By investigating a larger series of patients when\ncompared to the previous study, we also found a cutoff value that increases the\nchance of telling individuals with endometriosis from controls (PRL >17.5ng/mL).\nThe strength of the association was clearly confirmed in effect size analysis.\nThis finding agrees the observations made by  Bilibio\n et al . (2014) , which showed that serum prolactin\ncould also be applied as a test for peritoneal endometriosis. Some authors described\nthat increases in baseline serum prolactin might have a causative role in\ninfertility affecting patients with severe endometriosis ( Lima  et al ., 2006 ). It is unclear whether\nincreased prolactin levels are the cause or consequence of endometriosis. In fact,\nestradiol stimulates prolactin receptors in the uterus. In the presence of ectopic\nendometriotic tissues, prolactin receptors are overly induced. Inversely, increased\nbaseline serum prolactin reduces estrogen activity ( Gunin  et al ., 2002 ). Lowering prolactin secretion\nreestablishes functional ovulation and improves endometrial development. Other\nauthors have supported the use of prolactin inhibitors such as dopaminergic drugs to\nfavor fecundity ( Weil, 1986 ;  Crosignani, 2012 ). Others reported the use of\nantiestrogens such as Tamoxifen to decrease estrogen-stimulated prolactin levels in\nhyperprolactinemic rats ( Spritzer  et\nal ., 1996 ;  Aquino  et\nal ., 2016 ). Future studies might provide a better\nunderstanding of the role of estrogen-dependent medicine on the progression of\nendometriosis.\nIt was somewhat surprising to see that the prolactin level difference observed\nbetween individuals with disease stages I/II and stages III/IV in this study was not\nsignificant. The related effect size showed that the strength of association was\npoor. It seems plausible that the related non-significant  p -value\nmight be due to the inadequate size of the samples of individuals with different\nstages of endometriosis. The authors wondered whether the association might have\nbeen stronger if a larger sample had been selected. In future studies, it is\nsuggested that the associations between the various stages of endometriosis be\ninvestigated with larger samples of individuals with different stages of\nendometriosis.\nUnfortunately, we had trouble selecting individuals with endometriosis stage I.\nEndometriosis is often undiagnosed or misdiagnosed in affected women who come to the\nclinic looking for care. In other words, there is a time gap between the onset of\nsymptoms and the diagnosis of endometriosis.  Barbieri\n(2017)  reported a gap of more than eight years between the age of pelvic\nsymptom onset and the age of diagnosis. Possible explanations for this gap include\nlack of knowledge, variations in the manifestations of endometriosis, overlapping\nsymptoms with other pelvic diseases, unwillingness to undergo laparoscopy in early\nstage disease, concerns around nonsurgical methods, and the costs associated with\nendometriosis care. Nevertheless, most patients with minimal or mild endometriosis\nhave normal function and do not have to see a physician for early diagnosis. Most\nreturn for care after the disease has progressed. At the same time, affected women\nare exposed to higher blood prolactin levels for years. Even when not associated\nwith endometriosis, elevated prolactin levels produce devastating short- and\nlong-term effects and dramatically interfere with the reproductive and endocrine\nsystems of patients who are not treated in a timely manner ( Ballard  et al ., 2008 ). It is likely that\nimprovements to endometriosis care might shorten the time gap until diagnosis,\nparticularly in early-stage disease ( Weintraub,\n2016 ).\n\nInfertile women with more advanced endometriosis have higher prolactin levels than\ninfertile women without endometriosis. Prolactin is a probable prognostic biomarker\nto detect endometriosis stages III/IV vs. I/II and to differentiate infertile women\nwith endometriosis from infertile women without the condition. Prolactin levels\nmight be helpful in the detection of endometriosis.","source_license":"CC-BY-4.0","license_restricted":false}