Follicular fluid concentration of soluble Human-G Leukocytic Antigen (sHLA-G) in in vitro fertilization cycles of women with and without peritoneal endometriosis

Endometriosis is characterised by the presence of endometrial glands and stroma outside the endometrial cavity and the uterine musculature. While its pathogenesis remains unclear, this disease has been extensively studied in recent years and is one of the leading causes of the impairment of female fertility. Moreover, assisted reproduction techniques (ART) are one of the methods capable of reversing infertility in women with endometriosis. Recently, some investigators described the same reproductive results in patients with endometriosis submitted to IVF compared to non-endometriotic patients, despite the number of related abnormalities in those infertile patients with endometriosis described in the literature ( Giudice, 2010 ; Opøien et al. 2012 ; Barbosa et al., 2014 ). Soluble human leukocyte antigen-G (sHLA-G) is a molecule described to be linked to immunological tolerance of the semi-allogeneic fetus at the maternal-fetal interface during pregnancy as well as the activation of NK cells and even embryo quality response ( Hunt et al ., 2000 ; Noci et al., 2005 ; Jee et al., 2011 ; Rizzo et al., 2011 ; Wunder et al., 2013 ). sHLA-G has an important role in the activation of natural killer (NK) cells, Moreover, some authors associated sHLA-G in patients with endometriosis (mainly, the more severe forms) and adenomyosis ( Hunt et al. , 2000 ; Wang et al., 2008 ; Rached et al., 2019 ). However, the literature is completely absent in terms of the role of sHLA-G in infertile patients with endometriosis and the concentration of this molecule in the follicular fluid during controlled ovarian stimulation for IVF in this group of women. Endometriosis is not associated with altered embryo quality. However, some recente papers described that granulosa (cumulus) cell function could be altered in these women with peritoneal endometriosis submitted to IVF ( De Conto et al., 2017 ; 2021 ; Caran et al., 2021 ). sHLA-G was also described as a marker for oocyte and embryo quality, and some authors linked this molecule to pregnancy rates ( Rebmann et al., 2010 ). Furthermore, follicular fluid is a crucial environment to analyse and describe the association of sHLA-G and some reproductive outcomes in infertile patients with endometriosis ( Wunder et al., 2013 ). Considering the link of endometriosis with granulosa cell dysfunction and the possible role of sHLA-G as a marker for oocyte quality, the rationale of this study is to investigate the association between the concentrations of sHLA-G detected in the follicular fluid of infertile patients with peritoneal endometriosis submitted to IVF compared to non-endometriotic patients (tubal factor) as the primary outcome and its effect in some important reproductive outcomes.

Both groups were comparable in terms of age, AMH and reproductive outcomes ( Table 1 ). Pregnancy rates were also not different between the groups ( p =0.945) as the embryo score per transfer or the median embryo obtained per patient ( Table 1 ). The association (Spearman test) between the following variables was analysed: sHLA-G and AMH, age, number of collected MII, number of follicles, number of embryos and embryo score. All analyses did not demonstrate statistical significance ( p >0.05). Moreover, the multivariable analysis reinforced that sHLA-G was not related to any reproductive variable or outcome (age, number of follicles, number of follicles > 17mm, pregnancy rates, AMH, FSH and number of mature oocytes). Clinical characteristics and reproductive outcomes of included patients. Variables showed as mean±standard deviation (parametric data a ) or median and 95%CI (non-parametric data b ). t-student test; Mann-Whitney; Chi-square test. Performing the Mann-Whitney-U statistical test, the amount of sHLA-G measured in the group of women with peritoneal endometriosis and women without endometriosis was analysed, and we found no significant difference ( p =0.682). We also compared s-HLAG per follicle and per embryo between both groups of patients, and all analyses were not significant ( p =0.682, 0.955 and 0.857, respectively) ( Figure 1 ).

We demonstrated that peritoneal endometriosis was not associated with an altered sHLA-G follicular fluid concentration. Moreover, sHLA-G was not related to embryo quality, pregnancy rates, AMH, age or endometriosis. Endometriosis is a heterogeneous disease associated with infertility and one of the main causes of IVF. Furthermore, several mechanisms involving granulosa cells were described for this group of patients, involving prolactin secretion, decreased anti-Mullerian hormone and dysfunction in BMP-6 and SMAD4 in cumulus cells linked to peritoneal endometriosis ( Cunha-Filho et al. , 2001 ; 2003 ; De Conto et al., 2017 ; 2020). Nowadays the effect of endometriosis during an IVF cycle is disputed and has recently been questioned by several authors, but it is essential to understand the role of this disease during the IVF process to enhance pregnancy rates ( Opøien et al., 2012 ; Barbosa et al., 2014 ; González-Comadran et al., 2017 ; Caran et al., 2021 ). Therefore, we decided to better understand the granulosa cell (follicular compartment) of patients with peritoneal endometriosis submitted to IVF, analysing sHLA-G as a marker for oocyte/embryo competence and development ( Rizzo et al., 2007 ; Jee et al., 2011 ; Wunder et al., 2013 ). We included only peritoneal endometriosis because this specific phenotype was already described ( Santulli et al., 2016 ) as the most linked to infertility compared to the other phenotypes of endometriosis (endometrioma/deep). In this study, we also analysed the association of sHLA-G with endometriosis and with several potential markers of oocyte development and reproductive outcomes. The fact that sHLA-G is not related to endometriosis comes following recent papers showing that endometriosis per se is not a detrimental factor for patients submitted to IVF ( Barbosa et al., 2014 ; González-Comadran et al., 2017 ). Furthermore, sHLA-G seems to be an independent factor in oocyte competence/embryo quality, as demonstrated in a multicenter study ( Rebmann et al., 2010 ). Also, previous studies that showed an association between HLA system and endometriosis focused on the peritoneal fluid compartment and the immunological role of HLA molecules in patients with endometriosis. Deep endometriosis is more linked to several immunological abnormalities, and HLA molecules may play a role in the activation of NK cells in this group of patients ( Bylińska et al., 2018 ; Rached et al., 2019 ; Ścieżyńska et al., 2019 ). We have demonstrated that the immunological profile of peritoneal endometriosis is different from other phenotypes (deep and ovarian endometriosis), which could partially explain our results ( D’Hooghe et al., 2001 ; Glitz et al., 2009 ; Andreoli et al., 2011 ; Carmona et al., 2012 ). Follicular fluid sHLA-G concentration was associated with oocyte competence and embryo fertilization but not with good-quality embryos ( Jee et al., 2011 ). Besides, when performing oocyte maturation in close contact with cumulus oophorus complex (COCs), these cells produced sHLA-G during the oocyte maturation process, and sHLA-G was not detected in supernatants in the culture of COCs with immature oocytes, suggesting that sHLA-G is part of the oocyte maturation process but is not the limiting factor in that process ( Rizzo et al., 2009 ). We included strict criteria for embryo quality ( Fisch et al., 2001 ) to improve our statistical power and understand the relationship between sHLA-G and embryo development. As mentioned before and proved by our multifactorial analysis, sHLA-G is an independent factor (not associated with other reproductive parameters) that has been linked to embryo competence and even pregnancy rates after IVF by one study ( Rebmann et al., 2010 ). However, these authors stressed that embryo morphology is better than sHLA-G to select embryos. However, another group of authors did not find any association between sHLA-G in pregnant or not pregnant women after IVF ( Wunder et al., 2013 ). Furthermore, our study, when correlating the average score of the generated embryos and the concentrations of sHLA-G in the follicular fluid, concluded that the analysis of sHLA-G in FF is a parameter that is not associated with embryonic quality. This result corroborates with some studies in the literature. Future work could potentially utilise artificial intelligence or neural networks related to sHLA-G to improve embryo selection and better classify/choose the best embryo for transfer ( Bormann et al., 2020 ; VerMilyea et al., 2020 ). Our study had several limitations. First, we included only peritoneal endometriosis. Next, some reproductive outcomes need more subjects to be analysed, and for some outcomes, our number included patients was limited (pregnancy rate, for example), we calculated our sample size based on follicular fluid sHLA-G. Besides, sHLA-G is an unpractical tool, supplemented by embryo score and recently the advent of artificial intelligence. We included only superficial (peritoneal) endometriosis to perform a homogeneous group and increase our external validation. The main objective of this study was to investigate sHLA-G. For this purpose, we calculated the sample size, and our research included a sufficient number of patients. However, for reproductive outcomes, we need to include more patients, since analysing reproductive outcomes typically requires systematic reviews or international databases ( Barbosa et al., 2014 ; González-Comadran et al., 2017 ). The usefulness of sHLA-G during daily life in a human reproduction clinic is limited and obsolete, but it still does not invalidate this utility in terms of a research protocol. In this study, when analysing the sHLA-G, a kit with a detection limit of 0.6 Units/ml was used and therefore offered a greater sensitivity than the ELISA Kits described in the literature. The increase in this limit is directly related to the long incubation time (16-20 hrs) of the antibodies in contact with the FF samples, making it possible to detect sHLA-G in all the analysed samples in our study. Note that the importance of ELISA sHLA-G accuracy was already stressed by others ( Dahl & Hviid, 2012 ). We, therefore, emphasise the need for a more sensitive ELISA test both for studies with measurements in FF and embryo cultures.

We conclude that the level in the follicular fluid of sHLA-G in patients with superficial endometriosis was not altered compared to tubal factor patients submitted to IVF. Thus, we accept the null hypothesis that the concentrations of sHLA-G in follicular fluid in women with and without endometriosis do not differ.

A cross-sectional study was carried out from November 2016 to April 2021 at the Human Reproduction Center. One hundred-nine patients seeking infertility treatment (IVF) for the first time were included. Eligibility criteria used in the study were age ≤ 40 years, presence of both ovaries, hormone levels of TSH, FSH and PRL within the reference values and indication of in vitro fertilization procedure due to tubal factor or endometriosis, previously diagnosed by video laparoscopy, which was done at least six months before IVF (peritoneal endometriosis confirmed by biopsy and cauterised). Exclusion criteria were: ovarian hyperstimulation syndrome in the evaluated cycle, autoimmune disease, polycystic ovary syndrome, early luteinisation, endometrioma and presence of blood after follicular fluid centrifugation. Moreover, male partner semen analysis should be normal during semen preparation for IVF. We considered 109 patients with eligibility criteria, 13 patients were excluded; finally, 96 patients were included in this research, as demonstrated in Figure 1 . We subdivided into two groups, according to the cause of infertility-that is, with peritoneal endometriosis (n=46) and without endometriosis (n=50) -to assess the association between sHLA-G levels and embryonic quality. Figure 1 Follicular Fluid concentration of sHLA-G (IU/ml): a) in peritoneal endometriosis vs control group (non-endometriosis), p =0.682; b)comparing per follicle sHLA-G concentration between both groups, p =0.955 and (c) per embryo sHLA-G concentration ( p =0.857), Mann- Whitney-U test. Follicular Fluid concentration of sHLA-G (IU/ml): a) in peritoneal endometriosis vs control group (non-endometriosis), p =0.682; b)comparing per follicle sHLA-G concentration between both groups, p =0.955 and (c) per embryo sHLA-G concentration ( p =0.857), Mann- Whitney-U test. The presence or absence of endometriosis was verified by laparoscopy and biopsy, performed in the last 6 months before in vitro fertilization, and all included patients were diagnosticated with peritoneal endometriosis phenotype according to the European Society of Human Reproduction and Embryology ( Kennedy et al., 2005 ). The control group was formed by patients with only tubal fator as the infertility etiology. All patients in this group underwent laparoscopy, male analysis, hormonal screening for thyroid and propactin, regular menstrual cycle (21-35 days), age ≤ 40 years and presence of both ovaries. Hormonal dosages TSH, FSH and PRL were requested for the patients to be evaluated before beginning the cycle of assisted reproduction and were determined by chemiluminescence immunoassay (Siemens Om-MA Immulite 2000, Munich, Germany). Serum levels of AMH were determined by and enzyme-linked immunosorbent assay (Beckman Coulter, Inc., Brea, CA, USA). These results were transcribed from the medical records, and only those whose serum measurement was performed on the third day of the cycle were accepted. All patients were submitted to ovarian stimulation according to standard protocols. The therapeutic regimen for oocyte stimulation consisted of a GnRH antagonist for pituitary suppression with the administration of recombinant follicle-stimulating hormone (recombinant FSH) for ovarian stimulation. Follicular growth was monitored by transvaginal ultrasound, and when at least three follicles reached a diameter ≥ 17mm, the administration of chorionic gonadotropin (hCG) was determined to induce ovulation. Transvaginal follicular aspiration was performed 36 hours after the administration of hCG under routine intravenous sedation. At the time of oocyte laboratory recovery, follicular fluid was collected from a pool of punctured follicles (we only included in this study follicles superior to 17 mm with MII oocytes) for each cycle of the respective patient, centrifuged for 5 minutes at 1300 rpm, and, in the absence of blood, stored at -20°C for posterior analysis. Conventional IVF (in vitro fertilization) was performed 3 hours post puncture. Embryo score (GES, Graduated Embryo Score) criteria were used for assessing Embryonic Quality and Calculating the Average Embryo Score Generated). The evaluation of embryonic quality was carried out in three stages, from which scores ranging from 0 to 100 points were obtained. An embryo with 100 points is classified according to this criterion with the best morphological quality ( Fisch et al., 2001 ). The first moment of evaluation occurs 16-18 hours after IVF. At this moment, fertilization is confirmed by visualising the female and male pro-nuclei (PNs) and the presence of 2 polar corpuscles in the perivitelline space. The second moment occurs 25-27 hours after IVF and evaluates the early cleavage (presence of two cells), and the third moment of evaluation occurs 64-67 hours after IVF. In these three moments, the following parameters were evaluated: cell division and the presence of fragmentation with an embryo with maximum quality presenting 6 to 8 blastomeres of uniform size and no fragmentation at this time. After classifying the scores for each embryo generated from each cycle, the arithmetic mean between these scores was performed. In this study, the quality of the embryo was assessed by the average embryo score (EMEG) whose calculation was performed as follows: EMEG = (sum of scores of N embryos generated)/N (numbers of embryos generated). The levels of sHLA-G in the follicular fluid were measured by the enzyme-linked immunosorbent assay (ELISA) technique using the ELISA Kit (RD194070100R, Bio Venda, Czech Republic). The kit with the anti-sHLA-G monoclonal antibody can identify and measure the soluble isoforms HLA-G1 quantitatively (by proteolytic cleavage) and HLA-G5. Following the recommendations of the kit manufacturer, the samples were not diluted. The 16-20 hour incubation period between the FF sample and the anti-HLA-G monoclonal antibodies is an advantage of the kit since the literature describes considerably shorter periods, and a more extended incubation period allows greater test sensitivity (with a detection limit of 0.6 Units/ml). The plates were read at 450 nm, and the concentration of each sample analysed was calculated by the ELISA (Biotek ELX 800) that builds a calibration curve formed between the absorbance of the sample (Y-axis) and the concentration of the calibrators and samples tested (X-axis). The measured HLA-G concentration value is given in units per mL. The Ethics Committee approved this study under the number: 26453514.1.0000.5327. STROBE guidelines for observational studies were utilised for this research ( von Elm et al., 2008 ). Data analysis was performed using chi-square or Fisher’s exact tests for categorical data. Continuous variables were compared with Student’s t-test for parametric data, and Mann-Whitney-U was used for non-parametric data. Multiple comparisons were made using linear regression to address potential confounders. Furthermore, for correlations, we utilised Spearmen/Pearson tests. Seventy patients were required to have an 80% chance of detecting, as significant at the 5% level, a decrease in the primary outcome measure from 4.35 in the control group to 3 in the experimental group ( Rizzo et al., 2007 ). We also analysed the s-HLAG for follicles (dividing sHLA-G/number of follicles) and sHLA-G per embryo (total number of embryo/sHLA-G). These analyses were performed using the statistical program Statistical Package for the Social Science (SPSS) 21.0, and the data analysis was considered statistically significant when p <0.05.