- Email: [email protected]
Glycodelin in endometrial flushing fluid and endometrial biopsies from infertile and fertile women
European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Glycodelin in endometrial flushing fluid and endometrial biopsies from infertile and fertile women Ursula Bentin-Ley a,b, Anette Lindhard c,d, Vibeke Ravn e, Henrik Islin f, Steen Sørensen f,* a
The Department of Obstetrics and Gynecology, Herlev Hospital, University of Copenhagen, Denmark Danish Fertility Clinic, Copenhagen, Denmark c The Fertility Clinic, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark d The Fertility Unit, Roskilde Hospital, University of Copenhagen, Denmark e The Department of Pathology, Herlev Hospital, University of Copenhagen, Denmark f The Department of Clinical Biochemistry, Hvidovre Hospital, University of Copenhagen, DK-2650 Hvidovre, Denmark b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 16 August 2010 Received in revised form 19 November 2010 Accepted 27 December 2010
Objective: To investigate in the natural cycle just before IVF, whether glycodelin levels in endometrial flushing fluid obtained days LH + 1 and LH + 7 can be used in predicting pregnancy in the following IVF cycle, and whether there are differences in women with tubal factor infertility compared to women with unexplained infertility and fertile controls. Study design: A prospective observational multicentre study of 21 fertile and 75 infertile women (25 showed abnormal tubes with no signs of hydrosalpinges, 18 had uni- or bi-lateral hydrosalpinges, 17 were salpingectomised because of hydrosalpinges, and 15 women had unexplained infertility). Endometrial flushing at days LH + 1 and LH + 7, endometrial biopsy, and blood sampling at day LH + 7 were performed before down-regulation for IVF. Glycodelin levels in endometrial flushing fluids (EFF), biopsies, and plasma samples were related to tubal pathology, endometrial dating and IVF outcome. Furthermore, total protein concentration was measured in EFF to investigate the influence of normal endometrial maturation on protein concentrations from days LH + 1 and LH + 7. Results: At day LH + 1, EFF glycodelin levels were higher in infertile women with abnormal tubes compared to fertile women, particularly in women conceiving after the following IVF. For women with unexplained infertility, a higher level at day LH + 1 was present only in women not conceiving after the following IVF. ROC curve analysis showed that at day LH + 1 EFF glycodelin levels had no predictive value for IVF outcome. At day LH + 7, glycodelin levels in endometrial flushing fluids and biopsies depended on endometrial dating. Conclusions: At day LH + 1, glycodelin concentration is increased in endometrial flushing fluid from infertile women with abnormal tubes compared to fertile controls without being a valuable predictor of subsequent pregnancy. At day LH + 7 the glycodelin level depends on endometrial dating. ß 2011 Elsevier Ireland Ltd. All rights reserved.
Keywords: Endometrial flushing Endometrium Glycodelin Histology Immunohistochemistry Infertility PP14
1. Introduction Endometrial glands are the main source of plasma glycodelin (PP14) in the luteal phase [1]. The contribution from non-uterine tissues is small, although it is present in the salpinx [2,3], the human ovary [4], hematopoietic cells [5], and mammary glands [6]. For review, see [7]. Plasma glycodelin increases from five days after the endogenous LH peak (day LH + 5), peaks around the onset of menstruation, and then returns to baseline during the next follicular phase [8,9]. In conception cycles, plasma glycodelin rises rapidly after
* Corresponding author. E-mail address: [email protected] (S. Sørensen). 0301-2115/$ – see front matter ß 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2010.12.040
implantation, reaching a maximum at 8–10 weeks of gestation, and declines subsequently [10]. It is reduced in pregnant women without ovarian function [10], and in conception cycles in the first trimester after treatment with gonadotrophin releasing hormone agonist (GnRH-a), and particularly after clomiphene citrate stimulation, reflecting an altered endometrial function in pregnancies after ovarian stimulation [11]. Glycodelin can be demonstrated by immunohistochemical staining from day LH + 3 in endometrial glands, but neither in the stroma nor in the surface epithelium [12]. It has been demonstrated in endometrial flushing fluid (EFF) from day LH + 6 [13], and thereafter the level rises rapidly. In the late luteal phase glycodelin concentrations in EFF are more than 100-fold higher compared to plasma levels [13]. Lower glycodelin concentrations were found in EFF from women with unexplained infertility on days LH + 10 and
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
LH + 12, compared to fertile controls [14], whereas there were no detectable differences in EFF levels on day LH + 7 or in plasma samples on days LH + 7, LH + 10, and LH + 12 [14]. EFF glycodelin concentrations are correlated to endometrial dating according to Noyes’ criteria, whereas plasma glycodelin is not [15]. Because fluctuations of glycodelin levels are much more pronounced in EFF than in plasma [13], we wanted to investigate whether EFF glycodelin concentration at days LH + 1 and LH + 7, in the cycle just before IVF, is changed in women with various tubal pathologies including hydrosalpinx compared to women with unexplained infertility and fertile controls, and whether it can be used in predicting pregnancy in the following IVF cycle.
61
Ovulation day was defined as day LH + 1, and day LH + 7 as the middle of the midluteal phase, i.e. the implantation window. 2.2. Endometrial flushing Endometrial flushing was performed twice in every woman (at days LH + 1 and LH + 7). A small 8F urethral catheter with two separate channels was introduced into the endometrial cavity. One ml sterile saline was introduced through one of the channels into the endometrial cavity and aspirated through the other channel after 15 s of equilibration. This was repeated four times. Fluids were collected, centrifuged, and the supernatant stored at 20 8C until analysis.
2. Materials and methods 2.3. Blood samples 2.1. Patient populations Fertile controls and infertile women attending four Danish public fertility clinics for IVF treatment were included in the study after informed signed consent. All fulfilled the following inclusion criteria: age 25–38 years, body mass index (BMI) (19– 26 kg/m2), regular periods (21–35 days, variation 3 days between cycles), <10 cigarettes daily, and no hormonal treatments in the last three cycles before examination, normal ovaries with no signs of endometriosis by ultrasound examination. A positive urinary LH test indicated ovulation in the investigated spontaneous cycle. Couples needing intracytoplasmatic sperm injection were not included, but couples using donor sperm were accepted. Seventy-five infertile women completed the study with a subsequent IVF treatment, including embryo transfer (ET), and 21 of the fertile controls completed the investigation in a spontaneous cycle. The local Ethics Committees approved the study (KA96160m). 2.1.1. The tubal factor group Sixty women had tubal factor infertility, 25 showing abnormal tubes by hysterosalpingography (HSG) or laparoscopy but no signs of hydrosalpinges, neither at vaginal US before or during the trial (group A), 18 women had uni- or bi-lateral hydrosalpinges either at laparoscopy and/or at HSG and completed IVF treatment without salpingectomy (group B), and 17 women had been salpingectomised because of hydrosalpinges (group C). In some of our calculations and figures these three groups are put together as the tubal factor group (A + B + C).
At day LH + 7, blood samples were taken for later measurement of glycodelin. The samples were allowed to clot, and serum stored at 20 8C until assayed. 2.4. Measurements of glycodelin and total protein in EFF Measurement of glycodelin was performed by radioimmunoassay as described earlier [16]. The detection limit was 5 mg/l and inter-assay imprecision was 7.5%. All samples were assessed in duplicate. To investigate the influence of normal endometrial maturation on protein concentrations from days LH + 1 and LH + 7, the total protein concentration was measured in EFF by the Lowry method [17] and expressed as mg/l. 2.5. Endometrial biopsies An endometrial biopsy was obtained from the uterine fundus on day LH + 7 using a suction curette (pipelle de Cornier). Histological evaluation was performed on H&E stained sections, [()TD$FIG]
2.1.2. The unexplained infertility group Fifteen women had unexplained infertility with normal plasma FSH (<10 IU/l on cycle days 2–5), LH, prolactin and TSH on cycle days 2–5, a normal HSG and no history of pelvic inflammatory disease (PID). Their male partners all had normal semen samples (group D). 2.1.3. The fertile reference group The 21 fertile women were either volunteers or attending the hospitals for sterilization (group E). Fertility was documented by a history of having delivered at least one child, no recurrent spontaneous abortions, and no history of PID. The most recent child had been delivered within the past 4 years. They had no pregnancies for the last 6 months, or breastfeeding, or use of intrauterine device (IUD) or hormonal contraception for the last 3 months before the investigation. Plasma FSH levels on cycle day 2– 5 were <10 U/l in all women. 2.1.4. Prediction of ovulation time All women were examined in a spontaneous cycle, the infertile group in the cycle before IVF treatment. Ovulation was proven by a positive urinary LH test in the morning urine (Conceive, Quidel).
Fig. 1. SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (left) and Western blot analysis (right) of 7 mg amniotic fluid proteins with polyclonal rabbit antiglycodelin immunoglobulins. Lane M, molecular mass markers; lane 1 with no dithiothreitol treatment (DDT), and lane 2 with DDT. A band with a molecular mass of about 29 kDa is demonstrated (arrow).
[()TD$FIG]
62
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Fig. 2. Level of glycodelin in EFF at days LH + 1 (a) and LH + 7 (b) related to the diagnosis groups A–E.
using the criteria of Noyes [18–21] as a 2-day span interval of days following ovulation, i.e. post-ovulatory day (POD), and POD 5–7 was considered to be in-phase [18]. 2.6. Immunohistochemistry Glycodelin was purified from amniotic fluid for immunization purposes. An immunoadsorbent with rabbit antibody against human amniotic fluid proteins was used as first step [22], followed by anion exchange chromatography and a sequence of immunoaffinity chromatographies. The purified protein had the N-terminal sequence described for glycodelin (met, asp, ile, pro, gln, thr, lys, gln, asp, leu) and was used to raise a polyclonal antibody against glycodelin in rabbits. The antibody against glycodelin was purified by a positive immunoaffinity chromatography on a column coupled with glycodelin and tested towards amniotic fluid proteins on a PVDF membrane after being electrophoretically transferred from a 16% SDS-PAGE. The membrane was blocked, probed with the polyclonal antibody against glycodelin, and developed by polyclonal swine anti-rabbit immunoglobulins conjugated with horseradish peroxidase. A band was demonstrated with a molecular mass of about 29 kDa (Fig. 1). In the endometrial biopsy, glycodelin was demonstrated by an indirect immunohistochemical technique on sections pre-treated by microwaving (EGTA buffer pH9.0) using the Envision + kit with DAB+ as chromogen (K4111, Dako, Denmark). The first layer
antibody was diluted 1:8000 and incubated at 20 8C for 30 min. Negative controls included (1) omission of the first layer antibody, and (2) replacement of the primary antibody with a rabbit polyclonal antibody to helicobacter. Staining was evaluated separately for the different cell types present. The proportion of epithelial cells stained was evaluated semi-quantitatively as follows: score 0 (no cells stained), score 1 (<10%), score 2 (10–49%), score 3 (50–89%), and score 4 (90%). Scorings were based on subjective estimates [23]. 2.7. IVF treatment All infertile women were treated in the long protocol, using GnRH-a for down-regulation from day LH + 7; standard IVF was performed afterwards. Positive hCG per embryo transfer (ET) was between 29% and 50% in the four infertility groups, and ongoing pregnancy rate/ET was between 24% and 47% [18]. 2.8. Statistical analysis Differences between groups were investigated by Mann– Whitney U-test, a probability <0.05 was considered significant. Correlation between parameters was investigated by Spearman’s rank correlation coefficient rho. The predictive ability of EFF glycodelin levels for ongoing pregnancy was evaluated by a receiver–operator characteristic (ROC) curve analysis, which visualizes the trade-off between high
[()TD$FIG]
Fig. 3. Natural cycle levels of glycodelin in EFF at days LH + 1 (a) and LH + 7 (b), in women, conceiving (+hCG) and not conceiving ( hCG) after IVF, classified to diagnosis groups (tubal factor = group A + B + C) and compared to fertile controls.
[()TD$FIG]
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Fig. 4. A receiver–operator characteristic (ROC) curve analysis of the predictive ability for conception by determination of EFF glycodelin at day LH + 1.
63
compared to the fertile group E (Fig. 2a), but not at day LH + 7 (Fig. 2b). By combination of all tubal factor groups (A + B + C), irrespective of conception, they still had a significantly higher glycodelin level in EFF than the fertile group at day LH + 1 (Fig. 3a); this difference was not observed at day LH + 7 (Fig. 3b). Almost all the fertile women had EFF-glycodelin <50 mg/l at day LH + 1 (Fig. 3a). For women with unexplained infertility, a higher level at day LH + 1 was present only in women not conceiving after the following IVF (Fig. 3a), whereas no difference was present at day LH + 7 (Fig. 3b). In ROC curve analysis the area under the curve was 0.58 (CI95% 0.44–0.72) showing poor accuracy in predicting ongoing pregnancy at day LH + 1 by EFF glycodelin concentration (Fig. 4). 3.2. The relation of protein and glycodelin levels in endometrial flushing fluid to endometrial dating
sensitivity and high specificity. Calculation of sensitivity and specificity was done using different cut-off values of glycodelin concentration in EFF. 3. Results 3.1. Glycodelin in endometrial flushing fluid in infertile groups and prediction of IVF outcome At day LH + 1 EFF glycodelin levels were significantly higher in groups A and B (tubal factor with and without hydrosalpinges)
We compared the impact of normal endometrial maturation on glycodelin and protein concentrations in EFF from days LH + 1 and LH + 7. The comparison was done for both infertile and fertile women. No difference was present at day LH + 1 (Fig. 5a and b), although the fertile group showed a trend towards higher glycodelin concentrations in the histologically retarded endometria (p = 0.07). The total protein content, however, was statistically significantly elevated at day LH + 1 in EFF from histologically retarded endometria (POD 4), compared to endometria being in-phase (POD 5–7). This
[()TD$FIG]
Fig. 5. EFF glycodelin at day LH + 1 related to endometrial dating at day LH + 7 in infertile groups (a) and fertile controls (b), and EFF protein at day LH + 1 related to endometrial dating at day LH + 7 in infertile (c) and fertile women (d).
[()TD$FIG]
64
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Fig. 6. EFF glycodelin related to endometrial dating at day LH + 7 in infertile groups (a) and fertile controls (b), and EFF protein related to endometrial dating at day LH + 7 in infertile (c) and fertile controls (d).
outcome was found in both infertile (p = 0.03) and fertile women (p = 0.01) (Fig. 5c and d). On day LH + 7, a statistically significantly lower glycodelin concentration (p < 0.001) was found in EFF from histologically retarded endometria (POD 4) compared to normally matured endometria (POD 5–7) in the infertile group (Fig. 6a), and a trend in the same direction was seen for fertile women (p = 0.09) (Fig. 6b). The EFF protein concentrations at day LH + 7 showed no statistically significant differences between groups.
[()TD$FIG]
3.3. Glycodelin in endometrial flushing fluid and in glandular epithelium The glycodelin antibody stained the cytoplasm of some of the epithelial cells of the glands in 64 of 83 evaluable specimens, as well as secretions in glandular lumens. Staining scores of the glandular epithelium were related to POD. The median (range) staining score was for POD 4: 1 (0–2); for POD 5–7: 2 (0–4), and for POD 8: 2–3 (0–3). The surface epithelium was also stained but this staining was less frequent. The immunohistochemical staining score for glycodelin and EFF glycodelin levels at day LH + 7 showed a positive correlation. A Mann–Whitney test between mutual dependent groups (e.g. score 0 infertile versus score 0 fertile) showed no statistically significant differences (p > 0.05) (Fig. 7). 3.4. Plasma glycodelin No differences were observed in plasma glycodelin levels between infertile and fertile women at day LH + 7, all values being between 12 and 33 mg/l (data not shown). 3.5. Correlation between glycodelin, estradiol, and progesterone
Fig. 7. Glycodelin in EFF correlated to the staining grade of glycodelin in endometrial biopsies at day LH + 7.
Measurements from day LH + 7, comparing plasma progesterone, plasma estradiol, and plasma glycodelin with EFF glycodelin
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
65
Table 1 Correlation coefficient between the different parameters (Spearman rho). POD 4 Glycodelin in plasma (LH + 7) Tubal factor groupa Progesterone in plasma(LH + 7) Estradiol in plasma (LH + 7) Glycodelin in plasma (LH + 7) Unexplained infertility Progesterone in plasma (LH + 7) Estradiol in plasma (LH + 7) Glycodelin in plasma (LH + 7) Fertile controls Progesterone in plasma (LH + 7) Estradiol in plasma (LH + 7) Glycodelin in plasma (LH + 7)
POD 5–7 Glycodelin in EFF (LH + 1)
Glycodelin in EFF (LH + 7)
Glycodelin in plasma (LH + 7)
Glycodelin in EFF (LH + 1)
Glycodelin in EFF (LH + 7)
–
0.02 0.26 0.15
0.38 0.44* 0.35
0.15 0.02 –
0.43 0.25 0.15
0.14 0.04 0.003
Too few data Too few data –
Too few data Too few data Too few data
Too few data Too few data Too few data
0.61 0.05 –
0.29 0.13 0.86*
0.21 0.19 0.81*
0.1 0.2 0.4
0.1 0.2 0.1
0.13 0.33 0.14
0.003 0.13 0.54**
0.24 0.11
0.8 0.4 –
0.03 0.28 –
a
Groups A, B and C: abnormal tubes with no signs of hydrosalpinges, unilateral or bilateral hydrosalpinges, salpingectomised because of hydrosalpinges. * p = 0.03. ** p = 0.05.
showed a significant correlation between glycodelin in plasma and glycodelin in EFF in the unexplained infertile group and the fertile group, and between estradiol in plasma and glycodelin in EFF in those with tubal factor infertility (Table 1). 4. Comments Endometrial flushing with small amounts of physiological saline can be used for assessment of the secretory activity and function of the receptive endometrium. In this study, glycodelin was detected in EFF already at day LH + 1 and in a recent publication in endometrial secretion at POD 2 [24], which is somewhat earlier than in other investigations in which glycodelin was first detected from day LH + 5 [15]. In normal fertile women, the concentration rises rapidly in EFF after day LH + 6 and throughout the remaining cycle [13]. We found a higher concentration of glycodelin in EFF at day LH + 1 in women from groups A and B (tubal factor and persistent hydrosalpinges) compared to fertile controls, whereas salpingectomised and unexplained infertile women had levels comparable to the fertile controls (Fig. 2a). Normally, absence of the contraceptive glycodelin in the periovulatory uterus is consistent with a fertile window [25]. The reason for the increased periovulatory flushing glycodelin level in infertile women with persistent tubes is unknown, but difficulties in recovering the total volume of each flushing fluid may result in differences between blocked and patent salpinges [26]. Distally closed tubes leaking glycodelin into the uterine cavity might be another explanation. We did find a trend towards a lower level of EFF glycodelin in salpingectomised women (group C) at day LH + 1 compared to women with persistent pathological tubes. At day LH + 7 there were no differences between the groups. Abnormal tubes may cause impaired ovarian function, due to damaged blood supply, resulting in dysfunction of the endometrium. Impaired ovarian function may influence the glycodelin production since a positive correlation exists between plasma estradiol and glycodelin expression in endometrial biopsies [27], which has also been confirmed by endometrial gene expression studies [28]. In the present study, a positive correlation between plasma estradiol and EFF glycodelin at day LH + 7 was found in the tubal factor group. Synthesis and secretion of glycodelin also seem to be stimulated by progesterone [29], most likely through the Sp1 site in the glycodelin promoter region [30]. Glycodelin is negatively correlated to progesterone receptor B (PRB) in endometrial biopsies [31]. In our previous work, logistic regression analysis of multiple variables showed that the progesterone levels
at day LH + 7 were statistically significant for pregnancy outcome [18]. Since glycodelin changes during the luteal phase are much more pronounced in EFF than in plasma, we investigated the predictive value of EFF glycodelin at day LH + 1 for pregnancy in the coming IVF cycle by an ROC curve. This has to our knowledge not been done before. EFF glycodelin had no predicting value. There seems to be a clear relationship between EFF glycodelin concentration and endometrial maturation [15], which was also observed in the present study. By immunohistochemical staining of endometrial glands glycodelin has been shown to be present from day LH + 3 [12]. When endometria were out of phase, a significantly lower glycodelin expression was found in the late luteal phase (LH + 10 and LH + 13) compared to in-phase endometria [32]. A histologically retarded endometrium results in lower glycodelin content and may explain our findings at day LH + 7 (Fig. 6a). In fact, 65% of the biopsies in group A showed delayed endometrial development (POD 4) at day LH + 7, compared to 24% in the fertile reference group [18]. Otherwise, increased glycodelin expression has been shown in endometria during controlled ovarian hyperstimulation (COH), concordant with the advancement of endometrial maturation during COH [12]. Therefore, apart from endometrial development, it seems important to differentiate between natural cycles and COH cycles when evaluating glycodelin measurements in both biopsies and flushing fluids. The predictive value of plasma glycodelin has been investigated in several studies, reporting diverging results [14,33]. Comparison between plasma glycodelin levels in fertile and unexplained infertile women, not conceiving after IVF, showed no differences on days LH + 7, LH + 10 and LH + 12 [14]. Neither in our study could day LH + 7 plasma glycodelin predict pregnancy in the following IVF-treatment. It is important to find a method for investigating endometrial receptivity just prior to embryo transfer. This will enable us either to perform embryo transfer or to postpone transfer, freeze the embryos and transfer them in a natural cycle with a more receptive endometrium. The method must fulfill the demands of not traumatizing the endometrium prior to transfer. Endometrial flushing is probably not preferable. An aspiration of endometrial secretion and measuring glycodelin may give a good indication of endometrial maturation [24]. In conclusion, at day LH + 1 but not at day LH + 7, EFF glycodelin levels are higher in infertile women with abnormal tubes compared to fertile controls, but it is not a valuable predictor of pregnancy in a subsequent IVF cycle. In the midluteal phase, EFF glycodelin is correlated to endometrial maturation.
66
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Disclosure of interests The authors have nothing to disclose. The study sponsors had no role in the collection, analysis or interpretation of these data, or in the manuscript writing. Acknowledgements This work was funded by research grants from Organon A/S and Master Joiner Sophus Jacobsen and his wife Astrid Jacobsen’s Foundation. We wish to thank Vibeke Myrhøj for technical assistance and Susanne Bangsbøll, Sven Rex, Bente Toft and colleagues at the participating fertility clinics for their great work in obtaining the samples. References [1] Julkunen M, Koistinen R, Sjoberg J, et al. Secretory endometrium synthesizes placental protein 14. Endocrinology 1986;118:1782–6. [2] Julkunen M, Wahlstrom T, Seppala M. Human fallopian tube contains placental protein 14. Am J Obstet Gynecol 1986;154:1076–9. [3] Laird SM, Hill CJ, Warren MA, Tuckerman EM, Li TC. The production of placental protein 14 by human uterine tubal epithelial cells in culture. Hum Reprod 1995;10:1346–51. [4] Kamarainen M, Leivo I, Koistinen R, et al. Normal human ovary and ovarian tumors express glycodelin, a glycoprotein with immunosuppressive and contraceptive properties. Am J Pathol 1996;148:1435–43. [5] Kamarainen M, Riittinen L, Seppala M, Palotie A, Andersson LC. Progesteroneassociated endometrial protein—a constitutive marker of human erythroid precursors. Blood 1994;84:467–73. [6] Kamarainen M, Seppala M, Virtanen I, Andersson LC. Expression of glycodelin in MCF-7 breast cancer cells induces differentiation into organized acinar epithelium. Lab Invest 1997;77:565–73. [7] Seppala M, Taylor RN, Koistinen H, Koistinen R, Milgrom E. Glycodelin: a major lipocalin protein of the reproductive axis with diverse actions in cell recognition and differentiation. Endocr Rev 2002;23:401–30. [8] Joshi SG, Bank JF, Henriques ES, Makarachi A, Matties G. Serum levels of a progestagen-associated endometrial protein during the menstrual cycle and pregnancy. J Clin Endocrinol Metab 1982;55:642–8. [9] Westergaard LG, Wiberg N, Andersen CY, et al. Circulating concentrations of placenta protein 14 during the natural menstrual cycle in women significantly reflect endometrial receptivity to implantation and pregnancy during successive assisted reproduction cycles. Hum Reprod 1998;13:2612–9. [10] Johnson MR, Brooks A, Norman-Taylor JQ, et al. Serum placental protein 14 concentrations in the first trimester of ovum donation pregnancies. Hum Reprod 1993;8:485–7. [11] Johnson MR, Abbas A, Norman-Taylor JQ, et al. Circulating placental protein 14: in the first trimester of spontaneous and IVF pregnancies. Hum Reprod 1993;8:323–6. [12] Brown SE, Mandelin E, Oehninger S, et al. Endometrial glycodelin—a expression in the luteal phase of stimulated ovarian cycles. Fertil Steril 2000;74:130– 3. [13] Li TC, Ling E, Dalton C, Bolton AE, Cooke ID. Concentration of endometrial protein PP14 in uterine flushings throughout the menstrual cycle in normal, fertile women. Br J Obstet Gynaecol 1993;100:460–4.
[14] Mackenna A, Li TC, Dalton C, Bolton A, Cooke I. Placental protein 14 levels in uterine flushing and plasma of women with unexplained infertility. Fertil Steril 1993;59:577–82. [15] Li TC, Dalton C, Hunjan KS, Warren MA, Bolton AE. The correlation of placental protein 14 concentrations in uterine flushing and endometrial morphology in the peri-implantation period. Hum Reprod 1993;8:1923–7. [16] Ruge S, Sorensen S, Vejtorp M, Vejerslev LO. The secretory endometrial protein, placental protein 14, in women with ectopic gestation. Fertil Steril 1992;57:102–6. [17] Lowry O, Rosebrough N, Farr A, Randall R. Protein measurement with the Folin phenol reagent. J Biol Chem 1951;193:265–75. [18] Lindhard A, Ravn V, Bentin-Ley U, et al. Ultrasound characteristics and histological dating of the endometrium in a natural cycle in infertile women compared to fertile controls. Fertil Steril 2006;86:1344–55. [19] Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950;1:3–25. [20] Hendrickson MR, Kempson RL. Decision tree for endometrial dating. In: Benningtoon JL, editor. Surgical pathology of the uterine corpus. Philadelphia: WB Saunders; 1980. p. 80–5. [21] Dallenbach-Hellweg G. The endometrium of infertility. A review. Pathol Res Pract 1984;178:527–37. [22] Drohse H, Christensen H, Myrhoj V, Sorensen S. Characterisation of nonmaternal serum proteins in amniotic fluid at weeks 16–18 of gestation. Clin Chim Acta 1998;276:109–20. [23] Ravn V, Rasmussen BB, Hojholt L, et al. Reproducibility of subjective immunohistochemical estrogen- and progesterone receptor determination in human endometrium. Pathol Res Pract 1993;189:1015–22. [24] van der Gaast MH, Macklon NS, Beier-Hellwig K, et al. The feasibility of a less invasive method to assess endometrial maturation—comparison of simultaneously obtained uterine secretion and tissue biopsy. BJOG 2009;116:304–12. [25] Oehninger S, Coddington CC, Hodgen GD, Seppala M. Factors affecting fertilization: endometrial placental protein 14 reduces the capacity of human spermatozoa to bind to the human zona pellucida. Fertil Steril 1995;63: 377–83. [26] Hamilton JA, Iles RK, Gunn LK, et al. Concentrations of placental protein 14 in uterine flushings from infertile women: validation of the collection technique and method of expression of results. Hum Reprod 1998;13:3357–62. [27] Seppala M, Martikainen H, Ronnberg L, Riittinen L, Kauppila A. Suppression of prolactin secretion during ovarian hyperstimulation is followed by elevated serum levels of endometrial protein PP14 in the late luteal phase. Hum Reprod 1989;4:389–91. [28] Horcajadas JA, Pellicer A, Simon C. Wide genomic analysis of human endometrial receptivity: new times, new opportunities. Hum Reprod Update 2007;13:77–86. [29] Taylor RN, Vigne JL, Zhang P, et al. Effects of progestins and relaxin on glycodelin gene expression in human endometrial cells. Am J Obstet Gynecol 2000;182:841–7. [30] Gao J, Mazella J, Seppala M, Tseng L. Ligand activated hPR modulates the glycodelin promoter activity through the Sp1 sites in human endometrial adenocarcinoma cells. Mol Cell Endocrinol 2001;176:97–102. [31] Stavreus-Evers A, Mandelin E, Koistinen R, et al. Glycodelin is present in pinopodes of receptive-phase human endometrium and is associated with down-regulation of progesterone receptor B. Fertil Steril 2006;85:1803–11. [32] Klentzeris LD, Bulmer JN, Seppala M, et al. Placental protein 14 in cycles with normal and retarded endometrial differentiation. Hum Reprod 1994;9:394–8. [33] Andersen CY, Westergaard LG, Teisner B, et al. Changes induced in serum protein profiles by ovarian stimulation during in-vitro fertilization—embryo transfer treatment: a comparison between conception and non-conception cycles. Hum Reprod 1992;7:585–91.
Contents lists available at ScienceDirect
European Journal of Obstetrics & Gynecology and Reproductive Biology journal homepage: www.elsevier.com/locate/ejogrb
Glycodelin in endometrial flushing fluid and endometrial biopsies from infertile and fertile women Ursula Bentin-Ley a,b, Anette Lindhard c,d, Vibeke Ravn e, Henrik Islin f, Steen Sørensen f,* a
The Department of Obstetrics and Gynecology, Herlev Hospital, University of Copenhagen, Denmark Danish Fertility Clinic, Copenhagen, Denmark c The Fertility Clinic, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark d The Fertility Unit, Roskilde Hospital, University of Copenhagen, Denmark e The Department of Pathology, Herlev Hospital, University of Copenhagen, Denmark f The Department of Clinical Biochemistry, Hvidovre Hospital, University of Copenhagen, DK-2650 Hvidovre, Denmark b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 16 August 2010 Received in revised form 19 November 2010 Accepted 27 December 2010
Objective: To investigate in the natural cycle just before IVF, whether glycodelin levels in endometrial flushing fluid obtained days LH + 1 and LH + 7 can be used in predicting pregnancy in the following IVF cycle, and whether there are differences in women with tubal factor infertility compared to women with unexplained infertility and fertile controls. Study design: A prospective observational multicentre study of 21 fertile and 75 infertile women (25 showed abnormal tubes with no signs of hydrosalpinges, 18 had uni- or bi-lateral hydrosalpinges, 17 were salpingectomised because of hydrosalpinges, and 15 women had unexplained infertility). Endometrial flushing at days LH + 1 and LH + 7, endometrial biopsy, and blood sampling at day LH + 7 were performed before down-regulation for IVF. Glycodelin levels in endometrial flushing fluids (EFF), biopsies, and plasma samples were related to tubal pathology, endometrial dating and IVF outcome. Furthermore, total protein concentration was measured in EFF to investigate the influence of normal endometrial maturation on protein concentrations from days LH + 1 and LH + 7. Results: At day LH + 1, EFF glycodelin levels were higher in infertile women with abnormal tubes compared to fertile women, particularly in women conceiving after the following IVF. For women with unexplained infertility, a higher level at day LH + 1 was present only in women not conceiving after the following IVF. ROC curve analysis showed that at day LH + 1 EFF glycodelin levels had no predictive value for IVF outcome. At day LH + 7, glycodelin levels in endometrial flushing fluids and biopsies depended on endometrial dating. Conclusions: At day LH + 1, glycodelin concentration is increased in endometrial flushing fluid from infertile women with abnormal tubes compared to fertile controls without being a valuable predictor of subsequent pregnancy. At day LH + 7 the glycodelin level depends on endometrial dating. ß 2011 Elsevier Ireland Ltd. All rights reserved.
Keywords: Endometrial flushing Endometrium Glycodelin Histology Immunohistochemistry Infertility PP14
1. Introduction Endometrial glands are the main source of plasma glycodelin (PP14) in the luteal phase [1]. The contribution from non-uterine tissues is small, although it is present in the salpinx [2,3], the human ovary [4], hematopoietic cells [5], and mammary glands [6]. For review, see [7]. Plasma glycodelin increases from five days after the endogenous LH peak (day LH + 5), peaks around the onset of menstruation, and then returns to baseline during the next follicular phase [8,9]. In conception cycles, plasma glycodelin rises rapidly after
* Corresponding author. E-mail address: [email protected] (S. Sørensen). 0301-2115/$ – see front matter ß 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ejogrb.2010.12.040
implantation, reaching a maximum at 8–10 weeks of gestation, and declines subsequently [10]. It is reduced in pregnant women without ovarian function [10], and in conception cycles in the first trimester after treatment with gonadotrophin releasing hormone agonist (GnRH-a), and particularly after clomiphene citrate stimulation, reflecting an altered endometrial function in pregnancies after ovarian stimulation [11]. Glycodelin can be demonstrated by immunohistochemical staining from day LH + 3 in endometrial glands, but neither in the stroma nor in the surface epithelium [12]. It has been demonstrated in endometrial flushing fluid (EFF) from day LH + 6 [13], and thereafter the level rises rapidly. In the late luteal phase glycodelin concentrations in EFF are more than 100-fold higher compared to plasma levels [13]. Lower glycodelin concentrations were found in EFF from women with unexplained infertility on days LH + 10 and
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
LH + 12, compared to fertile controls [14], whereas there were no detectable differences in EFF levels on day LH + 7 or in plasma samples on days LH + 7, LH + 10, and LH + 12 [14]. EFF glycodelin concentrations are correlated to endometrial dating according to Noyes’ criteria, whereas plasma glycodelin is not [15]. Because fluctuations of glycodelin levels are much more pronounced in EFF than in plasma [13], we wanted to investigate whether EFF glycodelin concentration at days LH + 1 and LH + 7, in the cycle just before IVF, is changed in women with various tubal pathologies including hydrosalpinx compared to women with unexplained infertility and fertile controls, and whether it can be used in predicting pregnancy in the following IVF cycle.
61
Ovulation day was defined as day LH + 1, and day LH + 7 as the middle of the midluteal phase, i.e. the implantation window. 2.2. Endometrial flushing Endometrial flushing was performed twice in every woman (at days LH + 1 and LH + 7). A small 8F urethral catheter with two separate channels was introduced into the endometrial cavity. One ml sterile saline was introduced through one of the channels into the endometrial cavity and aspirated through the other channel after 15 s of equilibration. This was repeated four times. Fluids were collected, centrifuged, and the supernatant stored at 20 8C until analysis.
2. Materials and methods 2.3. Blood samples 2.1. Patient populations Fertile controls and infertile women attending four Danish public fertility clinics for IVF treatment were included in the study after informed signed consent. All fulfilled the following inclusion criteria: age 25–38 years, body mass index (BMI) (19– 26 kg/m2), regular periods (21–35 days, variation 3 days between cycles), <10 cigarettes daily, and no hormonal treatments in the last three cycles before examination, normal ovaries with no signs of endometriosis by ultrasound examination. A positive urinary LH test indicated ovulation in the investigated spontaneous cycle. Couples needing intracytoplasmatic sperm injection were not included, but couples using donor sperm were accepted. Seventy-five infertile women completed the study with a subsequent IVF treatment, including embryo transfer (ET), and 21 of the fertile controls completed the investigation in a spontaneous cycle. The local Ethics Committees approved the study (KA96160m). 2.1.1. The tubal factor group Sixty women had tubal factor infertility, 25 showing abnormal tubes by hysterosalpingography (HSG) or laparoscopy but no signs of hydrosalpinges, neither at vaginal US before or during the trial (group A), 18 women had uni- or bi-lateral hydrosalpinges either at laparoscopy and/or at HSG and completed IVF treatment without salpingectomy (group B), and 17 women had been salpingectomised because of hydrosalpinges (group C). In some of our calculations and figures these three groups are put together as the tubal factor group (A + B + C).
At day LH + 7, blood samples were taken for later measurement of glycodelin. The samples were allowed to clot, and serum stored at 20 8C until assayed. 2.4. Measurements of glycodelin and total protein in EFF Measurement of glycodelin was performed by radioimmunoassay as described earlier [16]. The detection limit was 5 mg/l and inter-assay imprecision was 7.5%. All samples were assessed in duplicate. To investigate the influence of normal endometrial maturation on protein concentrations from days LH + 1 and LH + 7, the total protein concentration was measured in EFF by the Lowry method [17] and expressed as mg/l. 2.5. Endometrial biopsies An endometrial biopsy was obtained from the uterine fundus on day LH + 7 using a suction curette (pipelle de Cornier). Histological evaluation was performed on H&E stained sections, [()TD$FIG]
2.1.2. The unexplained infertility group Fifteen women had unexplained infertility with normal plasma FSH (<10 IU/l on cycle days 2–5), LH, prolactin and TSH on cycle days 2–5, a normal HSG and no history of pelvic inflammatory disease (PID). Their male partners all had normal semen samples (group D). 2.1.3. The fertile reference group The 21 fertile women were either volunteers or attending the hospitals for sterilization (group E). Fertility was documented by a history of having delivered at least one child, no recurrent spontaneous abortions, and no history of PID. The most recent child had been delivered within the past 4 years. They had no pregnancies for the last 6 months, or breastfeeding, or use of intrauterine device (IUD) or hormonal contraception for the last 3 months before the investigation. Plasma FSH levels on cycle day 2– 5 were <10 U/l in all women. 2.1.4. Prediction of ovulation time All women were examined in a spontaneous cycle, the infertile group in the cycle before IVF treatment. Ovulation was proven by a positive urinary LH test in the morning urine (Conceive, Quidel).
Fig. 1. SDS-polyacrylamide gel electrophoresis (SDS-PAGE) (left) and Western blot analysis (right) of 7 mg amniotic fluid proteins with polyclonal rabbit antiglycodelin immunoglobulins. Lane M, molecular mass markers; lane 1 with no dithiothreitol treatment (DDT), and lane 2 with DDT. A band with a molecular mass of about 29 kDa is demonstrated (arrow).
[()TD$FIG]
62
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Fig. 2. Level of glycodelin in EFF at days LH + 1 (a) and LH + 7 (b) related to the diagnosis groups A–E.
using the criteria of Noyes [18–21] as a 2-day span interval of days following ovulation, i.e. post-ovulatory day (POD), and POD 5–7 was considered to be in-phase [18]. 2.6. Immunohistochemistry Glycodelin was purified from amniotic fluid for immunization purposes. An immunoadsorbent with rabbit antibody against human amniotic fluid proteins was used as first step [22], followed by anion exchange chromatography and a sequence of immunoaffinity chromatographies. The purified protein had the N-terminal sequence described for glycodelin (met, asp, ile, pro, gln, thr, lys, gln, asp, leu) and was used to raise a polyclonal antibody against glycodelin in rabbits. The antibody against glycodelin was purified by a positive immunoaffinity chromatography on a column coupled with glycodelin and tested towards amniotic fluid proteins on a PVDF membrane after being electrophoretically transferred from a 16% SDS-PAGE. The membrane was blocked, probed with the polyclonal antibody against glycodelin, and developed by polyclonal swine anti-rabbit immunoglobulins conjugated with horseradish peroxidase. A band was demonstrated with a molecular mass of about 29 kDa (Fig. 1). In the endometrial biopsy, glycodelin was demonstrated by an indirect immunohistochemical technique on sections pre-treated by microwaving (EGTA buffer pH9.0) using the Envision + kit with DAB+ as chromogen (K4111, Dako, Denmark). The first layer
antibody was diluted 1:8000 and incubated at 20 8C for 30 min. Negative controls included (1) omission of the first layer antibody, and (2) replacement of the primary antibody with a rabbit polyclonal antibody to helicobacter. Staining was evaluated separately for the different cell types present. The proportion of epithelial cells stained was evaluated semi-quantitatively as follows: score 0 (no cells stained), score 1 (<10%), score 2 (10–49%), score 3 (50–89%), and score 4 (90%). Scorings were based on subjective estimates [23]. 2.7. IVF treatment All infertile women were treated in the long protocol, using GnRH-a for down-regulation from day LH + 7; standard IVF was performed afterwards. Positive hCG per embryo transfer (ET) was between 29% and 50% in the four infertility groups, and ongoing pregnancy rate/ET was between 24% and 47% [18]. 2.8. Statistical analysis Differences between groups were investigated by Mann– Whitney U-test, a probability <0.05 was considered significant. Correlation between parameters was investigated by Spearman’s rank correlation coefficient rho. The predictive ability of EFF glycodelin levels for ongoing pregnancy was evaluated by a receiver–operator characteristic (ROC) curve analysis, which visualizes the trade-off between high
[()TD$FIG]
Fig. 3. Natural cycle levels of glycodelin in EFF at days LH + 1 (a) and LH + 7 (b), in women, conceiving (+hCG) and not conceiving ( hCG) after IVF, classified to diagnosis groups (tubal factor = group A + B + C) and compared to fertile controls.
[()TD$FIG]
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Fig. 4. A receiver–operator characteristic (ROC) curve analysis of the predictive ability for conception by determination of EFF glycodelin at day LH + 1.
63
compared to the fertile group E (Fig. 2a), but not at day LH + 7 (Fig. 2b). By combination of all tubal factor groups (A + B + C), irrespective of conception, they still had a significantly higher glycodelin level in EFF than the fertile group at day LH + 1 (Fig. 3a); this difference was not observed at day LH + 7 (Fig. 3b). Almost all the fertile women had EFF-glycodelin <50 mg/l at day LH + 1 (Fig. 3a). For women with unexplained infertility, a higher level at day LH + 1 was present only in women not conceiving after the following IVF (Fig. 3a), whereas no difference was present at day LH + 7 (Fig. 3b). In ROC curve analysis the area under the curve was 0.58 (CI95% 0.44–0.72) showing poor accuracy in predicting ongoing pregnancy at day LH + 1 by EFF glycodelin concentration (Fig. 4). 3.2. The relation of protein and glycodelin levels in endometrial flushing fluid to endometrial dating
sensitivity and high specificity. Calculation of sensitivity and specificity was done using different cut-off values of glycodelin concentration in EFF. 3. Results 3.1. Glycodelin in endometrial flushing fluid in infertile groups and prediction of IVF outcome At day LH + 1 EFF glycodelin levels were significantly higher in groups A and B (tubal factor with and without hydrosalpinges)
We compared the impact of normal endometrial maturation on glycodelin and protein concentrations in EFF from days LH + 1 and LH + 7. The comparison was done for both infertile and fertile women. No difference was present at day LH + 1 (Fig. 5a and b), although the fertile group showed a trend towards higher glycodelin concentrations in the histologically retarded endometria (p = 0.07). The total protein content, however, was statistically significantly elevated at day LH + 1 in EFF from histologically retarded endometria (POD 4), compared to endometria being in-phase (POD 5–7). This
[()TD$FIG]
Fig. 5. EFF glycodelin at day LH + 1 related to endometrial dating at day LH + 7 in infertile groups (a) and fertile controls (b), and EFF protein at day LH + 1 related to endometrial dating at day LH + 7 in infertile (c) and fertile women (d).
[()TD$FIG]
64
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Fig. 6. EFF glycodelin related to endometrial dating at day LH + 7 in infertile groups (a) and fertile controls (b), and EFF protein related to endometrial dating at day LH + 7 in infertile (c) and fertile controls (d).
outcome was found in both infertile (p = 0.03) and fertile women (p = 0.01) (Fig. 5c and d). On day LH + 7, a statistically significantly lower glycodelin concentration (p < 0.001) was found in EFF from histologically retarded endometria (POD 4) compared to normally matured endometria (POD 5–7) in the infertile group (Fig. 6a), and a trend in the same direction was seen for fertile women (p = 0.09) (Fig. 6b). The EFF protein concentrations at day LH + 7 showed no statistically significant differences between groups.
[()TD$FIG]
3.3. Glycodelin in endometrial flushing fluid and in glandular epithelium The glycodelin antibody stained the cytoplasm of some of the epithelial cells of the glands in 64 of 83 evaluable specimens, as well as secretions in glandular lumens. Staining scores of the glandular epithelium were related to POD. The median (range) staining score was for POD 4: 1 (0–2); for POD 5–7: 2 (0–4), and for POD 8: 2–3 (0–3). The surface epithelium was also stained but this staining was less frequent. The immunohistochemical staining score for glycodelin and EFF glycodelin levels at day LH + 7 showed a positive correlation. A Mann–Whitney test between mutual dependent groups (e.g. score 0 infertile versus score 0 fertile) showed no statistically significant differences (p > 0.05) (Fig. 7). 3.4. Plasma glycodelin No differences were observed in plasma glycodelin levels between infertile and fertile women at day LH + 7, all values being between 12 and 33 mg/l (data not shown). 3.5. Correlation between glycodelin, estradiol, and progesterone
Fig. 7. Glycodelin in EFF correlated to the staining grade of glycodelin in endometrial biopsies at day LH + 7.
Measurements from day LH + 7, comparing plasma progesterone, plasma estradiol, and plasma glycodelin with EFF glycodelin
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
65
Table 1 Correlation coefficient between the different parameters (Spearman rho). POD 4 Glycodelin in plasma (LH + 7) Tubal factor groupa Progesterone in plasma(LH + 7) Estradiol in plasma (LH + 7) Glycodelin in plasma (LH + 7) Unexplained infertility Progesterone in plasma (LH + 7) Estradiol in plasma (LH + 7) Glycodelin in plasma (LH + 7) Fertile controls Progesterone in plasma (LH + 7) Estradiol in plasma (LH + 7) Glycodelin in plasma (LH + 7)
POD 5–7 Glycodelin in EFF (LH + 1)
Glycodelin in EFF (LH + 7)
Glycodelin in plasma (LH + 7)
Glycodelin in EFF (LH + 1)
Glycodelin in EFF (LH + 7)
–
0.02 0.26 0.15
0.38 0.44* 0.35
0.15 0.02 –
0.43 0.25 0.15
0.14 0.04 0.003
Too few data Too few data –
Too few data Too few data Too few data
Too few data Too few data Too few data
0.61 0.05 –
0.29 0.13 0.86*
0.21 0.19 0.81*
0.1 0.2 0.4
0.1 0.2 0.1
0.13 0.33 0.14
0.003 0.13 0.54**
0.24 0.11
0.8 0.4 –
0.03 0.28 –
a
Groups A, B and C: abnormal tubes with no signs of hydrosalpinges, unilateral or bilateral hydrosalpinges, salpingectomised because of hydrosalpinges. * p = 0.03. ** p = 0.05.
showed a significant correlation between glycodelin in plasma and glycodelin in EFF in the unexplained infertile group and the fertile group, and between estradiol in plasma and glycodelin in EFF in those with tubal factor infertility (Table 1). 4. Comments Endometrial flushing with small amounts of physiological saline can be used for assessment of the secretory activity and function of the receptive endometrium. In this study, glycodelin was detected in EFF already at day LH + 1 and in a recent publication in endometrial secretion at POD 2 [24], which is somewhat earlier than in other investigations in which glycodelin was first detected from day LH + 5 [15]. In normal fertile women, the concentration rises rapidly in EFF after day LH + 6 and throughout the remaining cycle [13]. We found a higher concentration of glycodelin in EFF at day LH + 1 in women from groups A and B (tubal factor and persistent hydrosalpinges) compared to fertile controls, whereas salpingectomised and unexplained infertile women had levels comparable to the fertile controls (Fig. 2a). Normally, absence of the contraceptive glycodelin in the periovulatory uterus is consistent with a fertile window [25]. The reason for the increased periovulatory flushing glycodelin level in infertile women with persistent tubes is unknown, but difficulties in recovering the total volume of each flushing fluid may result in differences between blocked and patent salpinges [26]. Distally closed tubes leaking glycodelin into the uterine cavity might be another explanation. We did find a trend towards a lower level of EFF glycodelin in salpingectomised women (group C) at day LH + 1 compared to women with persistent pathological tubes. At day LH + 7 there were no differences between the groups. Abnormal tubes may cause impaired ovarian function, due to damaged blood supply, resulting in dysfunction of the endometrium. Impaired ovarian function may influence the glycodelin production since a positive correlation exists between plasma estradiol and glycodelin expression in endometrial biopsies [27], which has also been confirmed by endometrial gene expression studies [28]. In the present study, a positive correlation between plasma estradiol and EFF glycodelin at day LH + 7 was found in the tubal factor group. Synthesis and secretion of glycodelin also seem to be stimulated by progesterone [29], most likely through the Sp1 site in the glycodelin promoter region [30]. Glycodelin is negatively correlated to progesterone receptor B (PRB) in endometrial biopsies [31]. In our previous work, logistic regression analysis of multiple variables showed that the progesterone levels
at day LH + 7 were statistically significant for pregnancy outcome [18]. Since glycodelin changes during the luteal phase are much more pronounced in EFF than in plasma, we investigated the predictive value of EFF glycodelin at day LH + 1 for pregnancy in the coming IVF cycle by an ROC curve. This has to our knowledge not been done before. EFF glycodelin had no predicting value. There seems to be a clear relationship between EFF glycodelin concentration and endometrial maturation [15], which was also observed in the present study. By immunohistochemical staining of endometrial glands glycodelin has been shown to be present from day LH + 3 [12]. When endometria were out of phase, a significantly lower glycodelin expression was found in the late luteal phase (LH + 10 and LH + 13) compared to in-phase endometria [32]. A histologically retarded endometrium results in lower glycodelin content and may explain our findings at day LH + 7 (Fig. 6a). In fact, 65% of the biopsies in group A showed delayed endometrial development (POD 4) at day LH + 7, compared to 24% in the fertile reference group [18]. Otherwise, increased glycodelin expression has been shown in endometria during controlled ovarian hyperstimulation (COH), concordant with the advancement of endometrial maturation during COH [12]. Therefore, apart from endometrial development, it seems important to differentiate between natural cycles and COH cycles when evaluating glycodelin measurements in both biopsies and flushing fluids. The predictive value of plasma glycodelin has been investigated in several studies, reporting diverging results [14,33]. Comparison between plasma glycodelin levels in fertile and unexplained infertile women, not conceiving after IVF, showed no differences on days LH + 7, LH + 10 and LH + 12 [14]. Neither in our study could day LH + 7 plasma glycodelin predict pregnancy in the following IVF-treatment. It is important to find a method for investigating endometrial receptivity just prior to embryo transfer. This will enable us either to perform embryo transfer or to postpone transfer, freeze the embryos and transfer them in a natural cycle with a more receptive endometrium. The method must fulfill the demands of not traumatizing the endometrium prior to transfer. Endometrial flushing is probably not preferable. An aspiration of endometrial secretion and measuring glycodelin may give a good indication of endometrial maturation [24]. In conclusion, at day LH + 1 but not at day LH + 7, EFF glycodelin levels are higher in infertile women with abnormal tubes compared to fertile controls, but it is not a valuable predictor of pregnancy in a subsequent IVF cycle. In the midluteal phase, EFF glycodelin is correlated to endometrial maturation.
66
U. Bentin-Ley et al. / European Journal of Obstetrics & Gynecology and Reproductive Biology 156 (2011) 60–66
Disclosure of interests The authors have nothing to disclose. The study sponsors had no role in the collection, analysis or interpretation of these data, or in the manuscript writing. Acknowledgements This work was funded by research grants from Organon A/S and Master Joiner Sophus Jacobsen and his wife Astrid Jacobsen’s Foundation. We wish to thank Vibeke Myrhøj for technical assistance and Susanne Bangsbøll, Sven Rex, Bente Toft and colleagues at the participating fertility clinics for their great work in obtaining the samples. References [1] Julkunen M, Koistinen R, Sjoberg J, et al. Secretory endometrium synthesizes placental protein 14. Endocrinology 1986;118:1782–6. [2] Julkunen M, Wahlstrom T, Seppala M. Human fallopian tube contains placental protein 14. Am J Obstet Gynecol 1986;154:1076–9. [3] Laird SM, Hill CJ, Warren MA, Tuckerman EM, Li TC. The production of placental protein 14 by human uterine tubal epithelial cells in culture. Hum Reprod 1995;10:1346–51. [4] Kamarainen M, Leivo I, Koistinen R, et al. Normal human ovary and ovarian tumors express glycodelin, a glycoprotein with immunosuppressive and contraceptive properties. Am J Pathol 1996;148:1435–43. [5] Kamarainen M, Riittinen L, Seppala M, Palotie A, Andersson LC. Progesteroneassociated endometrial protein—a constitutive marker of human erythroid precursors. Blood 1994;84:467–73. [6] Kamarainen M, Seppala M, Virtanen I, Andersson LC. Expression of glycodelin in MCF-7 breast cancer cells induces differentiation into organized acinar epithelium. Lab Invest 1997;77:565–73. [7] Seppala M, Taylor RN, Koistinen H, Koistinen R, Milgrom E. Glycodelin: a major lipocalin protein of the reproductive axis with diverse actions in cell recognition and differentiation. Endocr Rev 2002;23:401–30. [8] Joshi SG, Bank JF, Henriques ES, Makarachi A, Matties G. Serum levels of a progestagen-associated endometrial protein during the menstrual cycle and pregnancy. J Clin Endocrinol Metab 1982;55:642–8. [9] Westergaard LG, Wiberg N, Andersen CY, et al. Circulating concentrations of placenta protein 14 during the natural menstrual cycle in women significantly reflect endometrial receptivity to implantation and pregnancy during successive assisted reproduction cycles. Hum Reprod 1998;13:2612–9. [10] Johnson MR, Brooks A, Norman-Taylor JQ, et al. Serum placental protein 14 concentrations in the first trimester of ovum donation pregnancies. Hum Reprod 1993;8:485–7. [11] Johnson MR, Abbas A, Norman-Taylor JQ, et al. Circulating placental protein 14: in the first trimester of spontaneous and IVF pregnancies. Hum Reprod 1993;8:323–6. [12] Brown SE, Mandelin E, Oehninger S, et al. Endometrial glycodelin—a expression in the luteal phase of stimulated ovarian cycles. Fertil Steril 2000;74:130– 3. [13] Li TC, Ling E, Dalton C, Bolton AE, Cooke ID. Concentration of endometrial protein PP14 in uterine flushings throughout the menstrual cycle in normal, fertile women. Br J Obstet Gynaecol 1993;100:460–4.
[14] Mackenna A, Li TC, Dalton C, Bolton A, Cooke I. Placental protein 14 levels in uterine flushing and plasma of women with unexplained infertility. Fertil Steril 1993;59:577–82. [15] Li TC, Dalton C, Hunjan KS, Warren MA, Bolton AE. The correlation of placental protein 14 concentrations in uterine flushing and endometrial morphology in the peri-implantation period. Hum Reprod 1993;8:1923–7. [16] Ruge S, Sorensen S, Vejtorp M, Vejerslev LO. The secretory endometrial protein, placental protein 14, in women with ectopic gestation. Fertil Steril 1992;57:102–6. [17] Lowry O, Rosebrough N, Farr A, Randall R. Protein measurement with the Folin phenol reagent. J Biol Chem 1951;193:265–75. [18] Lindhard A, Ravn V, Bentin-Ley U, et al. Ultrasound characteristics and histological dating of the endometrium in a natural cycle in infertile women compared to fertile controls. Fertil Steril 2006;86:1344–55. [19] Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950;1:3–25. [20] Hendrickson MR, Kempson RL. Decision tree for endometrial dating. In: Benningtoon JL, editor. Surgical pathology of the uterine corpus. Philadelphia: WB Saunders; 1980. p. 80–5. [21] Dallenbach-Hellweg G. The endometrium of infertility. A review. Pathol Res Pract 1984;178:527–37. [22] Drohse H, Christensen H, Myrhoj V, Sorensen S. Characterisation of nonmaternal serum proteins in amniotic fluid at weeks 16–18 of gestation. Clin Chim Acta 1998;276:109–20. [23] Ravn V, Rasmussen BB, Hojholt L, et al. Reproducibility of subjective immunohistochemical estrogen- and progesterone receptor determination in human endometrium. Pathol Res Pract 1993;189:1015–22. [24] van der Gaast MH, Macklon NS, Beier-Hellwig K, et al. The feasibility of a less invasive method to assess endometrial maturation—comparison of simultaneously obtained uterine secretion and tissue biopsy. BJOG 2009;116:304–12. [25] Oehninger S, Coddington CC, Hodgen GD, Seppala M. Factors affecting fertilization: endometrial placental protein 14 reduces the capacity of human spermatozoa to bind to the human zona pellucida. Fertil Steril 1995;63: 377–83. [26] Hamilton JA, Iles RK, Gunn LK, et al. Concentrations of placental protein 14 in uterine flushings from infertile women: validation of the collection technique and method of expression of results. Hum Reprod 1998;13:3357–62. [27] Seppala M, Martikainen H, Ronnberg L, Riittinen L, Kauppila A. Suppression of prolactin secretion during ovarian hyperstimulation is followed by elevated serum levels of endometrial protein PP14 in the late luteal phase. Hum Reprod 1989;4:389–91. [28] Horcajadas JA, Pellicer A, Simon C. Wide genomic analysis of human endometrial receptivity: new times, new opportunities. Hum Reprod Update 2007;13:77–86. [29] Taylor RN, Vigne JL, Zhang P, et al. Effects of progestins and relaxin on glycodelin gene expression in human endometrial cells. Am J Obstet Gynecol 2000;182:841–7. [30] Gao J, Mazella J, Seppala M, Tseng L. Ligand activated hPR modulates the glycodelin promoter activity through the Sp1 sites in human endometrial adenocarcinoma cells. Mol Cell Endocrinol 2001;176:97–102. [31] Stavreus-Evers A, Mandelin E, Koistinen R, et al. Glycodelin is present in pinopodes of receptive-phase human endometrium and is associated with down-regulation of progesterone receptor B. Fertil Steril 2006;85:1803–11. [32] Klentzeris LD, Bulmer JN, Seppala M, et al. Placental protein 14 in cycles with normal and retarded endometrial differentiation. Hum Reprod 1994;9:394–8. [33] Andersen CY, Westergaard LG, Teisner B, et al. Changes induced in serum protein profiles by ovarian stimulation during in-vitro fertilization—embryo transfer treatment: a comparison between conception and non-conception cycles. Hum Reprod 1992;7:585–91.