{"paper_id":"5c42bb28-c4c4-4215-955b-6a91c57d006f","body_text":"Assisted reproductive techniques undergo great advances,\nincluding improved means of tissue culture, updated\ncriteria of embryonic selection and extended culture to\nblastocyst, leading to reach pregnancy rates up to 66% in\nthe selected patients ( 1 - 4 ). In this regards, many factors\ninvolved in the implantation are not yet fully understood\nand it seems that endometrium plays much more relevant\nrole than the other recognized factors ( 5 - 7 ).\nEndometrium is a unique tissue that undergoes monthly\ncyclical changes resulting in menstruation, proliferation,\nsecretion and decidualization under the influence of\novarian steroids. Endometrium contains a large variety\nof immunocompetent cells, natural killers (NKs), macrophages,\nT cells and neutrophils, whose composition and\ndensity fluctuates periodically ( 8 ). The cycle-dependent\nchanges in these subpopulations of leukocytes and their\nmediators probably play a crucial role in implantation.\nIn contrast, antibody-bearing B-lymphocytes and plasma\ncells are rarely found in endometrial tissue ( 9 ).\nChronic endometritis (CE) is defined as localized inflammation\nof the endometrial mucosa characterized by\nthe presence of edema, increased stromal cell density, dissociated\nmaturation between epithelial cells and stroma\nfibroblasts, as well as the presence of plasma cell infiltrate\nin the stroma ( 10 ). These changes at the level of endometrial\nmicroenvironment could affect endometrial receptivity\n( 11 ).\nCE is often asymptomatic or present with non-specific\nclinical symptoms, such as pelvic pain, dysfunctional\nuterine bleeding, dyspareunia, vaginal discharge, vaginitis,\nrecurrent cystitis and mild gastro-intestinal discomfort\n( 12 ). Nonspecific quality of the symptoms and\nimportance of performing endometrial biopsy to confirm\ndiagnosis makes it difficult to estimate the prevalence of\nthis condition.\nBased on endometrial biopsy of patients who subsequently\nunderwent hysterectomy with benign pathology,\nprevalence of CE is 10-11% of the general population\n( 13 ,  14 ), 3-10% of patients with abnormal uterine bleeding ( 15 ) and up to 72% of women with suspected pelvic inflammatory disease (PID) due to the sexually transmitted diseases (STDs) ( 16 ). As far as infertile patients are concerned, the prevalence varies greatly depending on the utilized biopsy method and investigated population. In a prospective study published by Cicinelli et al. ( 17 ) in a total number of 2190 diagnostic hysteroscopy with different indications, they found a prevalence of 20% (438 patients) with CE, among whom 37% were also infertile. However, Kasius et al. ( 18 ) reported a prevalence of only 2.8%, in a total of 678 women.\nCE can be due to the presence of foreign bodies or structural pathology of the endometrial cavity, such as the presence of intrauterine device (IUD), submucous myomas, polyps, retained products of conception, incomplete abortion or infectious agents. The most frequent infectious agents are common bacteria frequently found in the urogenital area such as Streptococcus (27%), E. coli (11%), Enterococcus faecalis (14%) and Ureaplasma urealyticum (11%) ( 19 ). The presence of Chlamydia trachomatis is only 2.7%, and Neisseria gonorrhoeae is practically undetectable as causative in CE ( 20 ). These findings are in line with the results of the PEACH study ( 21 ), showing that 60% of women with PID present non-gonococcal or Chlamydia infection.\nIn certain areas of the world, Mycobacterium tuberculosis is highly prevalent. It is considered as the main cause of infertility in 40-75% of cases, since it causes implantation failure due to alteration of the immune response at the endometrial level, hormonal alterations and release of antiphospholipid antibodies ( 22 ). Today, it is well accepted that the uterus is not a sterile cavity, and that presence of the microorganisms is not equal to inflammation ( 23 ).\nAsymptomatic presence of bacteria in the endometrial cavity in either transcervical samples, or cultures obtained in post-hysterectomy specimens, has been reported by several investigators ( 24 ,  25 ). More recently metagenomics, investigating hypervariable regions of the ribosomal 16S rRNA genes allow definition of genus order and species of bacteria, leading to confirm presence of up to 12 different bacterial types in up to 95% of endometrial biopsies performed in patients undergoing hysterectomy for non-cancer indications ( 26 ). As described by Espinoza et al. ( 27 ), accumulating evidences suggest that endometrium is continuously exposed to bacteria from the genital tract. Therefore, presence of pathology is also determined by interaction of the infectious agent with the endometrial microenvironment ( 28 ).\nConsidering that the published data has not yet been able to draw a firm conclusion in this regard, in this paper we aim to review the current pieces of evidence regarding diagnosis, impact on reproductive outcomes and management of CE.\nIn the normal endometrium, B lymphocytes are only located at the basal layer, representing less than 1% of the leukocyte population. Conversely, in CE a large population of B cells lymphocytes are present at the both basal layer of the endometrium and glandular epithelium, as well as in the lumen of endometrial glands ( 29 ). Recent data suggest that a lipopolysaccharide derived from E. coli is capable of inducing the in vitro expression of E-selectin, as an adhesin that promotes passage of B cells to the endothelium of endometrial microvascularization ( 30 ). In addition, E-selectin promotes expression of chemoattractant CXCL13, activating adhesion molecules of B cells and expression of CXCL1 at the glandular endometrium level ( 8 ). In this microenvironment, gram-negative bacteria within the endometrium induce an abnormal immune response with migration of circulating B lymphocytes to the endometrial stromal compartment ( 8 ). At the endometrial level, plasma cells of the stroma express multiple immunoglobulins (IgM, IgA1, IgA1, IgG1 and IgG2), while excess of these antibodies could negatively affect implantation of the embryo ( 31 ).\nIn a study performed by Di Pietro et al. ( 32 ), expression of the 25 genes encoding the proteins involved in\ninflammation, proliferation and apoptosis at endometrial was compared by real-time polymerase chain reaction (RT-PCR) during the implantation time window in\n16 women with hysteroscopic and histological diagnosis\nof CE and 10 healthy women without endometritis; the\nresults of this study suggested that endometrial expression of some genes is significantly altered. In particular,\nthey found up-regulated gene expression of insulin-like\ngrowth factor binding protein 1 (IGFBP1), B-cell CLL/\nLymphoma 2 (BCL2) and BCL2-associated X protein\n(BAX), while down-regulated gene expression of IL11,\nChemokine (C-C MOTIF) Ligand 4 (CCL4), insulin-like\ngrowth factor 1 (IGF1) and caspase 8 (CASP8). These altered gene expressions could affect, at least in part, the\nembryonic implantation and they also explained presence\nof endometrial hyperplastic lesions.\nIn CE, stromal cells secrete IGFBP1 protein during the\ndecidualization process, exerting a negative effect on the\nimplantation process and counteracting effect of IGF2.\nThus, an increase of IGFBP1 expression and reduction of\nIGF1 expression in CE may lead to unfavorable conditions for implantation and embryonic development ( 33 ).\nIn this scenario, CE can alter the production of cytokines, impair endometrial function and induce an abnormal pattern of the leukocyte population at the endometrial\nlevel, leading to altered secretion of paracrine factors involved in endometrial receptivity. As reported elsewhere,\ndecrease in IL11 production by epithelial and stromal\ncells may lead to dysregulation of trophoblastic invasion,\nassociating with infertility. Similarly, lower CCL4 activity in CE may lead to a decreased recruitment of NKs and\nmacrophages, accounting for the observed implantation\nfailure ( 34 ). In addition, downregulation of BCL2 and\nCASP8 ( 35 ), associated with upregulation of BAX ( 36 ),\ncauses endometrial cell resistance to apoptosis and disturb the correct process of implantation ( 35 ).\nDiagnosis of CE represents a challenge for the gynecologist. The clinical manifestations of CE such as pelvic\npain, vaginal discharge, dyspareunia and abnormal vaginal bleeding are non-specific, while about 25% of patients\nwith CE are asymptomatic ( 37 ). Moreover, the peripheral\nblood inflammation markers, such as C-reactive protein\n(CRP), leukocytosis, leptin and IL6 do not predict presence of CE ( 8 ).\nThe accepted gold standard for diagnosis of CE is presence of the plasma cells in endometrial tissue. However,\ntheir histological identification is sometimes hampered\nby the presence of mononuclear cell infiltration, mitosis\nand proliferation of stromal cells, plasmacytoid appearance of stromal cells (fibroblasts and mononuclear cells)\nor decidual transformation of the endometrium during\nlate secretory phase. Plasma cells are characterized by the\npresence of chromatin in the form of a clock face inside\nan eccentric nucleus with perinuclear halo ( Fig .1 ) ( 15 ).\n: Immunochemistry of chronic endometritis.  A.  Fragment of endometrial biopsy specimen showing glandular cell surface syndecan 1 immunoreactivity.\nPlasma cells are highlighted by syndecan 1 staining in the center\nof the picture (original magnification: ×400) and  B.  More detailed picture\nof plasma cell syndecan 1 immunoreactivity.\nSyndecan 1 is a proteoglycan of the transmembrane\nheparan sulfate type presenting on the surface of plasma\ncells and keratinocytes, while it is not expressed in mononuclear cells, lymphocytes or endometrial stromal cells. It\nis also known as CD138, facilitating detection of plasma\ncells and presence of CE, not affected by intra- and inter-observer variability ( 38 ).\nIt is recommended to include clinical immunohistochemistry and conventional pathology study to increase the\naccuracy of the CE diagnosis ( 15 ). Furthermore, it is important to obtain standardization of the current diagnostic\ntechniques, considering that depending on the dilution of\nSyndecan 1, diagnosis of CE might differ. For a dilution of\n1:1000, a prevalence of 2.8% was initially reported for CE\nin asymptomatic infertile women prior to  in vitro  fertilization (IVF) ( 39 ), which does not overlap with the prevalence\nof 30.3%, previously reported by Johnston-MacAnanny et\nal. ( 40 ), as well as the prevalence of 10% in the patients\nwith recurrent miscarriage ( 41 ). In addition, the menstrual\ncycle phase whereby the biopsy is performed and thickness\nof the biopsy have paramount importance: in particular, in\n15% of the samples during secretory phase, plasma cells\nare present only in the basal layer of the stroma, which will\nbe missed if not included in the biopsy. Finally, it is important to define number of the plasma cells required to\nestablish diagnosis of CE: although most authors believe\nthat there must be two or more plasma cells, the others recommend presence of five or more plasma cells in at least\none of the three sections of biopsy ( 40 ).\nHysteroscopy is a useful diagnostic modality in CE.\nUsual hysteroscopic findings for characteristic CE include presence of local or diffuse hyperemia, edema of\nthe stroma and presence of micropolyps (less than 1 mm\nin size,  Fig .2 ) ( 42 ).\nDifferent findings of chronic endometritis at the fluid hysteroscopy.\n A.  Endometrial surface is completely covered by micropolyps,  B.  Isolated\nmicropolys on the lateral wall of the cavity,  C.  Endometrial mucosa appears thick, edematous, diffuse hyperemic, with presence of micropolyps,\nand  D.  Detailed image of an endometrial micropolyp appearance.\nCicinelli et al. ( 42 ,  43 ) reported that presence of endometrial micropolyps at hysteroscopy suggests the existence of CE. Interestingly, they obtained a positive diagnostic correlation of 93.4% with the pathology findings,\nfollowing their criteria of hysteroscopic diagnosis. These\nfindings have been replicated by others ( 44 ) with 86.5%\ncorrelation of hysteroscopic with histological diagnosis.\nThe implantation consists of a physiological process\ninvolving mediators of inflammation such as leukocytes,\ncytokines, chemokines and other endometrial factors. All\nthese cells and their mediators play an essential role in the\nregulation of immunoresponse and growth of the trophoblast. The presence of CE can alter receptivity of the endometrium creating an inadequate microenvironment that\ninterferes with normal implantation. In particular, recent\ndata ( 8 ) suggests that the endometrium of one third of infertile patients, presenting with CE, expresses high level of\nestrogen receptor, progesterone and Ki-67 nuclear marker\nof cell proliferation in both epithelial cells and stroma,\nin addition to the increased expression of anti-apoptosis\ngenes such as BCL2 and BAX, all of which represent a\nproliferative phenotypic change of the endometrium even\nin the secretory phase. This increase in expression levels\nof estrogen and progesterone receptors was replicated by\nWu et al. ( 33 ), suggesting that CE modifies stromal cells\nby altering the function of these hormonal receptors.\nCE also modifies the pattern of uterine contractility in\nboth of the periovulatory and mid-luteal phases of menstrual cycle ( 45 ). Physiologically, in the proliferative\nphase, there is anterograde contractility from the fundus\nto the cervix which facilitates removal of menstrual debris, followed by periovulatory and the luteal phase when\nthere is predominance of retrograde contraction in the opposite direction, from the cervix to fundus, which favors\nmigration of the spermatozoa to the fallopian tubes. Conversely, during CE, there is 3.3 times lower occurrence of\nretrograde contractility of the fallopian tubes ( 46 ). This\n“altered peristalsis” induced by the presence of CE could\nimpair, at least in part, fertility and contribute to some of\nthe symptoms such as pelvic pain and dysmenorrhea.\nImpact of the CE presence in implantation is controversial, although many studies suggest a negative impact\non the endometrial receptivity of plasma cells as well as\nIgM, IgG and IgA alterations in genes encoding proteins\ninvolved in the inflammatory response, proliferation and\napoptosis.\nBouet et al. ( 47 ) reported a prospective observational\nstudy including 46 women with recurrent implantation\nfailure (RIF), defined as failure to achieve pregnancy after\ntransferring three good quality embryos in fresh or frozen\ncycle in women up to 35 years of age, or 4 embryos of\ngood quality in women over 35 years. In this study, the\nauthors excluded women with uterine cavity anomalies,\npresence of submucous myomas or endometrial polyp of\nmore than 5 mm, as well as the patient who were treated\nwith antibiotics within one month prior to biopsy or those\nwho had unexplained vaginal bleeding. Hysteroscopy\nwith endometrial biopsy was performed between days 6\nand 12 of the cycle. Diagnosis of CE was confirmed by\n1:100 immunohistochemical dilution, while they were\nconsidered positive with the presence of 5 or more plasma cells in 10 high power fields (×400). They found 14%\nprevalence of CE, with 80% correlation between hysteroscopic criteria and histological confirmation. Using a\nslightly different methodology, others ( 40 ) investigated\nretrospectively 33 women, defining implantation failure\nas the failure to achieve pregnancy after two cycles of IVF\nwith transfer of at least one good-quality embryo. They\nperformed endometrial biopsy and immunohistochemical\nstudy dilution of 1:100, considering it negative with the\npresence of less than one plasma cell, reporting a prevalence of 30.3% CE. In a larger cohort analysis, Cicinelli\net al. ( 48 ) included the patients who were younger than 40\nyears, normal responders at ovarian stimulation and normal karyotype, defining RIF after embryo transfer of at\nleast six good quality embryos in three or more previous\nIVF/intracytoplasmic sperm injection (IVF/ICSI) cycles.\nPatients with follicle stimulating hormone (FSH)>10 on\nday 3, body mass index (BMI)>30, endometriosis, history of abortion, steroid use, autoimmune disease, antiphospholipid syndrome, thrombophilia or presence of\nanti-spermatozoid antibodies were excluded from the\nstudy. They performed hysteroscopy and biopsy in the\nfollicular phase of the following cycle, obtaining an endometrial biopsy and cultured these cells. According to\ntheir data analysis, CE was diagnosed by hysteroscopy in\n66% of the cases, by histology in 57.5%, and by positive\nculture in 45% of the cases. Higher rate of positive diagnosis could be explained probably by the experience of\npathologist and hysteroscopist, regarding the diagnostic\ncriteria of CE ( 40 ) and a selection bias since the authors’\nInstitution is a referral center for women with suspected\nCE. The final concordance between hysteroscopic and\nhistologic diagnosis of CE was 87%. Noteworthy to say\nthat women who were included in both studies, performed\nby Johnston-MacAnanny et al. ( 40 ) and Cicinelli et al.\n( 48 ), were treated with antibiotic. Patients included in the\nformer study ( 40 ) were treated with 100 mg Doxycyclin\nfor two weeks, followed by Ciprofloxacin and Metronidazole 500 mg (twice daily) for two weeks in those with\npositive cultures. Those included in the latter study ( 48 )\nwere treated with Ciprofloxacin 500 mg (twice daily) for\n10 days against gram negative bacteria and AmoxicillinClavulanic acid 1 g (twice daily) for 8 days against gram\npositive bacteria. If the cultures persisted positive, then\nthe antibiotic protocol was repeated up to three times and\nif the cultures were negative, the patient would receive intramuscular single dose of Ceftriaxone 250 mg, followed\nby Doxycycline 100 mg (twice daily) and Metronidazole\n500 mg (twice daily) for 14 days.\nRegarding reproductive outcomes, Cicinelli et al. ( 48 )\nfound a live born rate of 61% in patients responding to\nantibiotics, whereas the live born rate was only 13% in\npatients who did not respond to antibiotic therapy. Conversely, in the study performed by Johnston-MacAnanny\net al. ( 40 ), patients of the CE group improved pregnancy\nrate after good response to therapy, although the CE group\nhad still lower pregnancy rate than non-CE group, despite\na good response to treatment with antibiotics. These different results may probably be due to the other unrecognized endometrial abnormalities, which are not solved\nwith antibiotic therapy.\nOverall, both studies suggest that CE has a negative\nimpact on endometrial receptivity, and adequate response\nto antibiotic therapy may significantly improve reproductive outcomes, as it was confirmed in a recent systematic review and meta-analysis ( 49 ). Nevertheless, diagnostic\nhysteroscopy itself and endometrial biopsy may also play\na positive role. In one hand, hysteroscopy could physically remove bacterial biofilms involved in the pathophysiology of CE; on the other hand, endometrial biopsy and the\nsubsequent recovery process can promote secretion of cytokines and growth factors in the endometrium involved\nin embryo implantation.\nAccording to European Society of Human Reproduction and Embryology (ESHRE), recurrent pregnancy loss\n(RPL) is defined as the loss of two or more pregnancy,\neven not consecutive, occurring before 20 weeks of gestation, which is in agreement with the definition of the\nAmerican Association of Reproductive Medicine (ASRM)\nguidelines ( 50 ). In patients with implantation failure, the\naberrant endometrial microenvironment resulting from an\nanomalous pattern in the CE lymphocyte population has\nbeen linked to RPL. Kitaya et al. ( 41 ) reported a total of\n58 women with RPL (three or more abortions), detecting\npresence of CE by immunohistochemistry in 9.3% of the\npatients. Using the same experiment, others ( 51 ) reported\na prevalence of 42.9% CE on a total of 142 women with\nthree or more abortions. McQueen et al. ( 52 ) studied 395\nwomen with two or more abortions by week 10 or at least\none pregnancy loss of more than 10 weeks, finding 9%\nCE prevalence diagnosed by endometrial biopsy. In the\nlatter study, the patients were then treated with antibiotics: after the first course, there was adequate response in\n94% of the cases, rising to 100% after administration of\ntwo courses of antibiotics. They reported an increase of\nlive birth rate from 7% before treatment to 56% after receiving antibiotic treatment for two weeks.\nCicinelli et al. ( 53 ) performed a retrospective study of\n360 women under the age of 40 with three or more abortions before 20 weeks gestation, excluding patients with\nsevere male factor, endometriosis, uterine anomalies,\nmetabolic or hormonal alterations, antiphospholipid syndrome and thrombophilia. Hysteroscopy was performed\nin the follicular phase. Patients with hysteroscopic diagnosis of CE had endometrial biopsy in the following cycle. They found 57.8% of patient with hysteroscopic sign\nof CE, out of which 91.3% were confirmed by histology\nand 68% had positive cell cultures. Confirming previous reports, after antibiotic treatment they found that live\nbirth rate in women responding to antibiotic treatment\nwas higher, compared to non-responder women, suggesting that presence of the infectious agents in the uterine\ncavity has a potential deleterious impact on the endometrial environment.\nSimilarly, a more recent case-control observational\nstudy ( 54 ) was performed in 107 women with two or\nmore abortions before 20 weeks gestation, after ruling out\nother causes of pregnancy loss. In this study, investigators\nperformed endometrial biopsy analyzed with hematoxylin eosin and CD138, defining CE as the presence of 1-5\nplasma cells at immunohistochemistry test. Using these\ncriteria, the prevalence of CE varied from 13% to 56%\nupon completion of an immunohistochemical study. They\nalso found a trend towards a higher rate of pregnancy loss\nin women with untreated CE compared to patients without CE. Finally, Bouet et al. ( 47 ) published a prospective observational study, including 53 women with two\nor more unexplained pregnancy loss in pregnancies less\nthan 14 weeks gestation. They performed hysteroscopy\nand endometrial biopsy, using syndecan 1, as a biomarker,\nconsidering that is positive with the presence of five or\nmore plasma cells in 10 high power fields. They found a\nprevalence of 27% CE.\n\nCE is associated with poor reproductive outcomes,\nincluding implantation failure and RPL. Accumulating\nevidences suggest that this condition modifies endometrial microenvironment at different levels: first of all, CE\npromotes changes on immunocompetent cell population\nin the endometrium. It also affects production of inflammatory cytokines, involved in NKs recruitment, which\nplay a crucial role in local immune response during early\npregnancy and favor implantation. In addition, CE has\nnegative impact on normal endometrial decidualization,\npromoting proliferation, diminishing apoptosis and modifying the expression of sex steroid receptors, which affect\nendometrial receptivity.\nHysteroscopy, in expert hands, could be considered a\ngood tool to combine with histology for diagnosis of CE.\nNevertheless, a consensus about strict criteria is mandatory for diagnosis to combine immunohistochemistry\nwith conventional histology. Finally, future investigation\nshould be aimed to redefine the minimum volume of biopsy and the number of plasma cells needed for diagnosis.\nThere is still lack of the uniform definition of RPL. Obtaining that would allow more accurate analysis and comparison among different studies. Considering this scenario, part of conflicting data found by different authors can\nbe due to this element.\nAntibiotic treatment of CE improves implantation rates\nand decreases the rate of abortion, although there is a lack\nof well-designed prospective studies that corroborate this\nfinding.\nThe metagenomics and a better understanding of the\nmicrobioma of the reproductive tract will allow researchers to develop therapies aimed to not only eliminate pathogenic flora but also establish a flora which favors reproductive success.","source_license":"CC-BY-4.0","license_restricted":false}