{"paper_id":"ecdd8638-83c0-402e-9b8d-7572af62af14","body_text":"Adenomyosis is a common gynecologic disease in premenopausal and late reproductive age women. Frequent symptoms include progressive dysmenorrhea, chronic pelvic pain, dyspareunia, abnormally heavy menstrual bleeding, and interval menstrual bleeding. The severity and frequency of the symptoms correlate with the extent and depth of the ectopic endometrium in the myometrium, which follows the endometriosis pathology and progression of the disease ( 1 ). Generally, diffuse globular enlargement is the common sign of adenomyosis and the size of uterus is typically not larger than 12 weeks gravid size. This enlargement of the uterus is due to invasion of the myometrium by endometrial glands only be made at histopathology following hysterectomy.\nCurrently, the etiology and pathology of adenomyosis\nhave yet to be elucidated, however, many theories have been\nproposed; for example, the invagination of basalis endometrium\ninto the myometrium and the subsequent inflammation.\nAdenomyosis uteri repetitively exhibit exaggerated\nand non-synchronized uterine contractions, which induce a\nmicro-fracture at the endometrial-myometrial junction zone\n(EMJZ). This micro-trauma at EMJZ causes the displacement\nof endometrium into the surrounding myometrium,\nwhere the myometrial cells proliferate and undergo metaplasia,\nresulting in the thickening of EMJZ ( 3 ,  4 ). The aberrant\nimmunologic activity in adenomyosis is seen by immunohistochemistry\nwith an increase in the number of infiltrating macrophages. These could activate T and B cells to produce\nantibodies and cytokines that may destroy the EMJZ ( 5 ).\nMacrophages transform from the circulatory monocyte cells\nand serve many functions including phagocytosis, antibodydependent\ncell-mediated cytotoxicity, and presenting antigens\nto lymphocytes. Natural killer (NK) cells, on the other\nhand, are cytotoxic lymphocytes that play a role in innate\nimmune response. They can remove infected cells without\nrecognition, and thus the killing process is an immediate response\nto virus-infected cells and tumor formation ( 6 ,  7 ).\nAbnormal immune responses appear to play a role in adenomyosis.\nExpression of human leukocyte antigen (HLA)\nclass II is recognized as the first step in the activation of\nmacrophages and/or T lymphocytes by presenting foreign\nantigens to these cells. HLA class II expression is shown to\nbe increased in the glandular cells of eutopic and ectopic\nendometrium of adenomyosis ( 6 ,  7 ). At this point, the definitive\ntreatment for adenomyosis is hysterectomy ( 5 ), but\nit is not the treatment of choice for patients who wish to\nremain fertile. Oral contraceptive pills, high doses of progestin,\nlevonorgestrel-releasing intrauterine device (LNGIUD),\ngonadotropin releasing hormone agonist (GnRH-a),\nand danazol are hormonal treatment options for both adenomyosis\nand endometriosis ( 1 ). The use of pre-operative\nGnRH agonists in patients with myomas, endometriosis\nand adenomyosis was tested by Khan et al. ( 8 ). They\nfound a decrease in MCP-1 level, as well as a reduction\nin the number of macrophages in endometrial and myometrial\nlayers of patients with adenomyosis compared to\nan untreated group. Progestin is considered a good choice\nfor long-term treatment of adenomyosis because it causes\nminimal side effects. The administration of a progestin, depending\non the dose, may not cause a hypoestrogenic state.\nFour generations of synthetic progestogens are now available to clinicians. A new progestin called dienogest (DNG) ( 17a-cyanomethyl-17b-hydroxyestra-4, 9-dien bioavailability and strong progestational effect due to its high selectivity to the progesterone receptor. It has been reported as one of the progestogens employing the strongest action on the endometrium ( 10 ). Its half-life is 9-10 hours, and it is approximately 90% secreted by the kidneys. The therapeutic dose of DNG for treatment of endometriosis is 2 mg daily providing a blood level of 10-7 mol/L. Although the blood level of dienogest after oral administration is very low, the symptoms of endometriosis can be cured ( 9 ). Dienogest has been shown to slightly increase the number and macrophages ( 11 ). The objective of our study was to investigate the effect of DNG on immune cells focusing on macrophages and NK cells in women with adenomyosis.\n\nA randomized control trial study was conducted from 1 February 2015 to 31 May 2016 at the Department of Obstetrics and Gynecology, Faculty of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand. The study was approved by the Ethical clearance Committee on human rights related to researches involving human subjects ( pro women aged 20-50 years, suffering from dysmenorrhea with or without hypermenorrhea, undergoing either abdominal hysterectomy or laparoscopic hysterectomy for adenomyosis, and diagnosed adenomyosis by ultrasonography. The exclusion criteria were patients who had used i. Pill or oral exogenous hormones for 3 months or ii. GnRH agonist or depot medroxyprogesterone acetate injection for 6 months before surgery. Twenty-four eligible patients were recruited into the study. The ultrasonographic diagnostic criteria for adenomyosis included a globular shape with diffuse uterine enlargement, myometrial anterior-posterior asymmetry, poorly described areas in the myometrium, a heterogeneous myometrial echotexture, sub-endometrial echogenic linear striations with or without small myometrial cysts, or hemorrhagic foci within the heterotopic endometrial layer ( 12 ). The sample size was calculated according to measurement outcomes and the mean number of NK cells and macrophages from Khan’s study with ple size was 12 for each arm. Therefore, a total of 24 subjects were needed for this present study.\nThe subjects were divided into two groups via block randomization\nwith a block size of 4. A group of 12 patients received\n2 mg/day dienogest for 28-35 days before surgery, and\nthe other 12 underwent operation without prescribing the drug\n(control group). The phase of the menstrual cycle was determined\nby the subjects’ last menstrual period and confirmed\nby endometrial histology. Patients with liver disease, kidney\ndisease, autoimmune disease, coagulopathy, endometriotic\ncysts, or under hormonal therapy were excluded. Exclusion of\npatients with pelvic endometriosis was not possible; therefore,\nit was the potential limitation for the present study.\nThe myometrial tissue had gross features of adenomyosis, that is, hypertrophic swirls of smooth muscle separating duller prominent trabeculated patterns and gray foci of endometrium. The tissue was excised in 1cmx1cmx1cm dimentions immediately after hysterectomy. The endometrial tissue sample was also collected in block pattern size 1 cm×1 cm×1 cm from the myometrial layer. All collected sections of the samples were then prepared for subsequent histopathological and immunohistochemical studies.\nImmunohistochemical analysis was performed with anti-\nCD68 antibody for macrophages and anti-CD57 and anti-\nCD56 for NK cells. Both primary rabbit monoclonal antibodies\nclone PGM-1 and clone 123C3.D5 against CD68\nand CD57, respectively, were from DAKO (Glostrup,\nDenmark), and the primary mouse monoclonal antibody\nclone NK-1 against CD56 was from Thermo Fisher Scientific (Waltham, USA). Primary antibodies CD68, CD57,\nand CD56 were used at a dilutions of 1:100, 1:150, and\n1:2400, respectively. The 3-μm thick paraffin-embedded\ntissue sections were deparaffinized in xylene and then rehydrated.\nSlides were incubated for 60 minutes at 60°C\nand treated with Bond Dewax Solution (Leica Biosystems,\nBannockburn, IL). Epitope retrieval was performed by\nincubating the slides in Bond Epitope Retrieval Solution\nfor 20 minutes at 100°C. Immunohistochemical analysis\nwas performed using the Bond Polymer Refine Detection\nkit (Leica Biosystems, Bannockburn, IL), a 3-step indirect\nimmunoperoxidase technique. Briefly, primary antibody\nwas applied for 45 minutes at room temperature. Peroxide\nblock (3% hydrogen peroxide) was then applied for\n5 minutes and rinsed with Bond Wash Solution. Post Primary\nPolymer was applied for 9 minutes. Polymer Poly-\nHRP IgG was applied for 7 minutes and rinsed with Bond\nWash Solution and deionized water, then diaminobenzidine\nchromogen was applied for 4 minutes. Slides were counterstained\nwith hematoxylin for 5 minutes. Appendix tissue\nwas used as positive control. The negative control showed\nan absence of specific staining ( 12 ).\nThe number of CD68 and CD57 brown spots were counted\nin 20 different fields (200×200 microns) for each person\n(magnification: ×200) under light microscopy. The number of\nCD68 and CD57 positive cells were calculated and expressed as the mean positive cells per mm  2.Theresultsineachbiopsy er who did not know the patient’s history. Double-labeling immunohistochemical method for CD57 and CD56 was also performed. The primary antibody anti-CD57 was the first antibody\nidentified by diaminobenzidine chromogen, while the primary antibody anti-CD56 was the second antibody identified\nby mixed red refine chromogen. The procedure was\nsimilar to the described protocol above including deparaffinization, epitope retrieval and a 3-step indirect immunoperoxidase technique. However, before counterstaining each slide with hematoxylin, the second antibody (anti-CD56 primary antibody) was applied for 40 minutes at room temperature. Post Primary Polymer AP was applied for 20 minutes. Polymer Poly-AP IgG was applied for 20 minutes and rinsed with Bond Wash Solution and deionized water before the Mixed were then counterstained with hematoxylin for 5 minutes.\nEach parameter is presented as either the mean ± SD or\nmedians (25, 75%) depending on the distribution of data. The\nclinical characteristics were compared by chi-square test or\nstudent’s t test for differences between two groups. The differences\nin numbers of macrophages and NK cells between\nthe two groups were analyzed by the non-parametric Mann-Whitney U-test; a value of P<0.05 was considered statistically\nsignificant. The data were analyzed by IBM SPSS Statistics\nfor Windows, version 19.0 (Armonk, NY: IBM Corp).\n\nTwenty-four elibible patients were enrolled and randomly\ndivided into treated and untreated groups. There\nwere no dropouts after treatment of dienogest for 28-35\ndays, therefore, complete data from 24 participants were\navailable for analysis ( Fig .1 ). The demographic characteristics\nof the participants; for example, age, BMI, indication\nfor and type of surgery were not significantly\ndifferent ( Table 1 ). The number of participants with proliferative\nand secretory menstrual phases in the DNG and\nuntreated groups were not statistically different.\nStudy design for the effects of dienogest on macrophages and natural\nkiller cells in adenomyosis clinical trials.\nDemographic data of women with adenomyosis\nBMI; Body mass index, AUB; Abnormal uterine bleeding, and NSAIDs; Nonsteroidal antiinflammatory\ndrug.\nNK cell infiltration, as shown by CD57-positive brown spots ( Fig .2 ), was significantly increased in glands in\ntreated versus untreated groups [P=0.005, median (range\n25, 75%), 4.37 (0.31, 14.68) vs. 0 (0, 0)] ( Fig .2D ), but\nnot in stroma of the endometrium of the treated group\n[P=0.416, median (range 25, 75%), 21.87 (10, 68.81), 10\n(8.75, 15.93)] ( Fig .2F ,  Table 2 ). NK cells were confirmed\nby double immunostaining for CD57 and CD56 (Fig.2G,\nH). Macrophage infiltration, as demonstrated by CD68-\npositive brown spots ( Fig .3 ), was also increased, but was\nnot statistically significant in treated versus untreated\ngroups [P=0.055, median (range 25, 75%), 4.37 (0, 9.06),\n0 (0, 1.56)] ( Fig .3D ,  Table 2 ).\nMacrophage and natural killer (NK) cell infiltration in endometrial glands and stroma in endometrial tissue in adenomyosis patients\nData as presented as [median (range 25-75%)].\nImmunohistochemistry staining for CD57 in eutopic and ectopic endometrium. The accumulation of CD57 is indicated by a\ndark brown coloration in the nucleus; all nuclei were counter-stained with hematoxylin.  A.  Ectopic endometrium in myometrium\nof adenomyosis women in untreated group,  B.  Ectopic endometrium in myometrium of adenomyosis women in dienogest-treated\ngroup,  C.  Endometrial gland in endometrium of untreated group,  D.  Endometrial gland in endometrium of dienogest-treated group,\n E.  Endometrial stroma in endometrium of untreated group,  F.  Endometrial stroma in endometrium of dienogest-treated group,  G. \nDouble staining for CD57 and CD56 in endometrium, and  H.  Double staining for CD57 and CD56 in myometrium.\nInfiltration of macrophages and NK cells in the myometrium\nof women with adenomyosis was also investigated. No\ndifferences were seen between DNG-treated and untreated\ngroups (Figes .2 ,  3 ,  Table 3 ).\nThe uterine endometrial thickness was more significantly\nreduced after DNG treatment for 30 ± 2.76 days\n(mean ± SD) when compared to the untreated group [median\n(25, 75%), 37.5 (33.37, 60.12), 94.5 (67.75, 143.25),\nrespectively] (P=0.021).\nMacrophage and natural killer (NK) cells infiltration in ectopic endometrium in myometrial tissue in adenomyosis patients\nData as presented as [median (range 25-75%)].\nImmunohistochemistry staining for CD68 in eutopic and ectopic endometrium. The accumulation of CD68 is indicated by a\ndark brown coloration in the nucleus and all nuclei were counter-stained with hematoxylin. A. Ectopic endometrium in myometrium\nof adenomyosis women in untreated group, B. Ectopic endometrium in myometrium of adenomyosis women in dienogest-treated\ngroup, C. Endometrial gland in endometrium of untreated group, D. Endometrial gland in endometrium of dienogest-treated group,\nE. Endometrial stroma in endometrium of untreated group, and F. Endometrial stroma in endometrium of in denogest-treated group.\nThis randomized controlled trial compared the numbers\nof infiltrating macrophages and NK cells in the eutopic\nand ectopic endometrium in women with adenomyosis\nafter either one month of treatment with DNG or no\ntreatment. The results showed that oral DNG increased\nthe number of NK cell infiltration in glandular structure\nof eutopic endometrium but not in ectopic endometrium.\nHowever, no differences were observed in the number of\nmacrophages and NK cells in stromal tissue of eutopic endometrium\nor ectopic endometrial tissue of the two study\ngroups. This suggests that progestin has a direct effect on\nendometrium in addition to its systemic effect, thus may\nimprove local immunity in the endometrium.\nThe histopathological characterization of adenomyosis\nis similar to endometriosis, except for the site of endometriotic\nlesions. Immune dysfunction and inflammation\nplay significant roles in both endometriosis and adenomyosis.\nVarious immune alterations occur in endometriosis\nthat are also present in adenomyosis. We noted an increase\nin the number and activity of macrophages secreting proinflammatory\ncytokines interleukin-1 (IL-1) and tumor\nnecrosis factor (TNF) in women with endometriosis. It\nhas been shown that patients with adenomyosis have both\nHLA-DR and HLA-G expression in eutopic and ectopic\nendometrium ( 7 ,  14 ). The expression of HLA-G corresponding\nto major histocompatibility complex (MHC)\nclass I increases the survival of endometrial cells from\nthe host’s immunosurveillance ( 13 ). Ota and Igarashi ( 7 )\nand Ota et al. ( 15 ) reported an increased number of macrophages\nin ectopic and eutopic endometrium in women\nwith adenomyosis. Recently, Zhihong et al. ( 16 ) did a prospective\ncase-control study by collecting the endometrial\ntissue from patients with adenomyosis during the implantation\nwindow after ovarian stimulation, and then performed\nimmunohistochemistry and real time-polymerase\nchain reaction (RT-PCR). The results showed an increase\nin cytokine IL-6, monocyte chemoattractant protein-1\n(MCP-1) and the number of macrophages. MCP-1, which\nis produced by uterine NK cells is a cytokine that induces\nthe migration of immune cells to target tissue.\nNK cells play a role in destroying viruses and tumor\ncells but not normal host cells. The killing activity depends\non the activity of killer cell inhibitory receptors\n(KIRs) and MHC molecules ( 17 ). The reduction of NK\ncell activity in peripheral blood and peritoneal fluid has\nbeen shown in endometriosis ( 18 ,  19 ). The NK cell cytotoxicity\nmay also be reduced in eutopic endometrium in\nadenomyosis, because the expression of CD94a (surface\nmarker for KIR) is increased in the endometrium versus\nthe myometrium in women with adenomyosis ( 17 ).\nThere are a few classical NK cells CD56+CD16+ in both\neutopic and ectopic endometrium in adenomyosis ( 20 ).\nHere, we used CD57-a terminally sulfated carbohydrate\nepitope (glucuronic acid 3-sulfate)-to identify NK cells.\nThis marker identifies the final stage of NK cell maturation.\nTherefore, these matured CD57+ NK cells may have\nstrong cytotoxic activity, but weak proliferative property\n( 21 ). Additionally, we verified NK cells via dual immunohistochemistry\nstaining for CD56 and CD57.\nSeveral previous studies have demonstrated progesterone\nand progestin-modulated immune responses ( 22 - 26 ).\nThe cyclical change of progestin in endometriums of nonpregnant\nwomen suggested hormone regulation, specifically\nprogesterone ( 27 ). Our results showed that DNG increased\nthe number of mature NK cells at the glandular\nepithelium of eutopic endometrium, which is comparable\nto Klinger’s findings. Oral combined contraceptive pills\ncontaining DNG have been demonstrated to increase the\nnumber of lymphocytes, monocytes and granulocytes in\npatients treated for one cycle ( 22 ). A similar result was\nseen in an animal study, where administration of DNG in\nrats with endometrial tissue auto-transplantation slightly\nincreased the NK cell activity and number in peritoneal\nfluid cells and spleen cells ( 11 ).\nProgestin treatment including megestrol acetate for endometrial\nhyperplasia and endometrial cancer-induced\nimmune suppression can increase the number and activity\nof NK cells and other immune cells ( 24 ). NK cells\nare the important immune cell type in the gestational\ndecidua. The regulation of uterine NK cells by progesterone\nand progestin are well-known during pregnancy\n( 27 ,  28 ). However, the role of uterine NK cells remains\nunclear. Many are present during implantation, and they\nmay be involved in the implantation mechanism ( 23 ). The\npotential role of uterine NK cells is that it supports the\npreparation of uterus for embryo implantation by producing\nmany cytokines. Moreover, Le Bouteiller and Piccinni\n( 27 ) found that uterine NK cells in early decidua could\nkill target cells, for example, infected maternal decidual\ncells, supporting the local immune responses to uterine\ninfection ( 23 ,  27 ,  29 ). Therefore, DNG administration in\npatients with adenomyosis may improve the implantation\nprocess and protect the fetus from infection to pregnancies\nthat occur after discontinuation of treatment.\nDifferent types of progestin may affect immune responses\ndifferently ( 23 ,  30 ). However, there is no direct\ncomparative study on the effect of different progestins\non NK cells. This would be an interesting topic for further\nstudies. Progesterone recruites uterine NK cells by\nincreasing the chemokine C-X-C motif ligand (CXCL)10\nand CXCL11, which has been demonstrated in an in vitro\nstudy using an endometrial organ culture system ( 31 ). In\naddition, hormone replacement therapy has been shown\nto change NK cell activity ( 32 ). In this study, DNG enhanced\nthe NK cell number only at the glandular structure\nof the endometrium. There was no increase in NK\ncell numbers in stromal tissue of eutopic endometrium or\nin the ectopic endometrium. In the study conducted by\nMehasseb et al. ( 33 ), they examined the expression pattern\nof progesterone receptor (PR)-A and PR-B and foci\nlesion of adenomyosis by immunohistochemistry in the\nendometrium of control and adenomyosis subjects. The\nexpression of both PRs was lower in the stroma, and the inner and outer myometrium in the adenomyotic samples\ncompared to glands. Therefore, the differential effect of\nDNG on various tissues may have been mediated by differential\nexpression of PRs on adenomyotic tissue and eutopic\nendometrium.\nDNG did not affect macrophage infiltration on eutopic\nand ectopic endometrium or on the myometrium from\npatients with adenomyosis. The effect of DNG on macrophage\ninfiltration was different from that of GnRHa\nshown by Khan et al. ( 8 ). GnRHa decreased the infiltration\nof CD68-expressing cells in the endometrium of\nwomen with endometriosis and adenomyosis. It is possible\nthat the local actions of DNG and GnRHa on endometriosis\nor adenomyosis lesions were different. The\nhighlight of our study is that, patients were randomly assigned\nto two study groups by block randomization with a\nblock size of 4. There was an equal number of participants\nin each study group and the cohorts had homogeneous\nfeatures. This maintained the balance of the study groups\nand reduced selection bias. In this study we had aimed to\ninvestigate the effect of progestin on mature NK cells by\nimmunohistochemistry, which indirectly reflect NK cell\ncytotoxicity. However, further studies on functional NK\ncells and other characteristics of NK cells, i.e. proliferation\nor apoptosis, are still needed.\n\nProgestin DNG administration causes an increase in\nuterine mature NK cells in glandular structure of eutopic\nendometriumn in patients with adenomyosis. The immunomodulating\neffect of progestin on adnenomyosis may\nbe beneficial for implantation and fetal protection to pregnancies\noccurring after treatment. The enhancing effect of\nprogestin on NK cells is differentially expressed depending\non the site of immune infiltration.","source_license":"CC0","license_restricted":false}