| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Departments of Gynecology and Obstetrics (N.F.V., C.W.L., B.B., T.-H.L., J.A.K., Y.Z.) and Pathology (I.-M.S.), Johns Hopkins University School of Medicine, Baltimore, and Bloomberg School of Public Health (K.F.), Baltimore, Maryland 21205; Department of Gynecology/Obstetrics (T.-H.L.), Cathay General Hospital, Taipei 106, Taiwan; Department of Medicine (T.-H.L.), School of Medicine, Fu Jen Catholic University, Taipai, Taiwan, Republic of China 24205; and Second Department of Obstetrics and Gynecology (N.F.V.), University of Athens School of Medicine, 11528 Athens, Greece
Address all correspondence and requests for reprints to: Yulian Zhao Ph.D., 601 North Caroline Street, JHOC 1211, Baltimore, Maryland 21287. E-mail: yzhao1{at}jhmi.edu.
| Abstract |
|---|
|
|
|---|
Objective: Our objective was to evaluate the impact of different luteal-phase support protocols on sex steroid levels and on endometrial expression of L-selectin ligand after ovarian hyperstimulation with a GnRH antagonist protocol.
Patients and Design: Seventeen oocyte donors who underwent ovarian stimulation with a recombinant FSH/ganirelix acetate protocol were randomized into three groups: group I had no luteal-phase support; group II had luteal support with micronized progesterone; and group III had luteal support with progesterone plus 17ß-estradiol. All donors had endometrial biopsies on the day of retrieval, and then 3, 5, and 10 d after retrieval. In addition, they had serum estradiol and progesterone measurements on d 3, 5, and 10.
Main Outcome Measures: Endometrial L-selectin ligand expression was detected by immunohistochemical staining in the luminal and glandular epithelium. A histological score was used for the quantification of the immunostaining. Sex steroid levels were measured during the luteal phase.
Results: By d 10 after retrieval, there was a significant decrease in mean progesterone levels in group I compared with the other two groups that may reflect the expected demise of the corpus luteum. There was also a significant increase in the presence of L-selectin ligands in the luminal epithelium in group III.
Conclusions: During controlled ovarian stimulation with a GnRH antagonist protocol, luteal-phase support with micronized progesterone and 17ß-estradiol seem to increase endometrial L-selectin ligand expression in the luminal endothelium.
| Introduction |
|---|
|
|
|---|
Before and during implantation, extensive cross-signaling occurs between the embryo and the endometrium. This involves a finely tuned spatial and temporal production and secretion of specific hormones, monokines, cytokines, and growth factor-binding proteins as well as adhesion molecules, some of which have been used as markers of endometrial receptivity (2, 6, 7, 8).
Recently, mechanisms involving L-selectin and L-selectin ligand interaction have been described in the process of human implantation (9). Selectins constitute a group of cell adhesion molecules that mediate transient cell to cell interactions necessary for the recirculation of lymphocytes from the blood to the lymphoid organs (homing) and back to the blood through postcapillary venules located in the cortex of the lymph nodes (10). These venules are characterized by the height of their endothelium, and they are now commonly called high endothelial venules (HEV). The molecule responsible of the adhesion of lymphocytes to the HEV is a type I membrane protein identified as L-selectin (CD62L) (11) that interacts with HEV-located L-selectin ligands. Identification of HEV-based L-selectin ligands has been achieved with a monoclonal antibody MECA-79, which has been shown to block lymphocyte attachment to HEV in the Stamper-Woodruff in vitro adherence assay and inhibit short-term lymphocyte homing (12). MECA-79 effectively blocks the tethering and rolling of lymphocytes along HEV, thus preventing the initiation of the recruitment cascade (13). MECA-79 and L-selectin-IgG chimeras immunoprecipitate the same complex of proteins from mouse lymph nodes and human tonsils, possibly because of the resemblance of its sulfated carbohydrate epitope with the L-selectin recognition determinant sialyl Lewis X (6-sulfo sLex) (14).
In the study by Genbacev et al. (9), endometrial biopsies obtained from oocyte donors revealed a strong staining of the luminal epithelium for L-selectin ligand during the luteal phase. Subsequent immunoblot analysis with MECA-79 confirmed an up-regulation of L-selectin ligands as the window of implantation opened (d 3 and 6 after retrieval compared with d 0 and 2). A similar pattern was observed in endometrial samples obtained from women during their natural cycle (15). Staining of embryos at different stages of development with a specific L-selectin antibody was weak, when the zona pellucida was intact, whereas there was a strong trophectoderm staining for L-selectin after hatching. Epithelial binding of cytotrophoblasts onto the receptive luteal-phase endometrium was effectively inhibited by adding an antibody to L-selectin or by preincubation with MECA-79 (9). These findings support a pivotal role for the L-selectin system in the process of implantation.
In the present study, we evaluate the impact of different types of luteal-phase support after ovarian stimulation with a GnRH antagonist protocol on the endometrial receptivity by using the expression of L-selectin ligands as a surrogate marker. The availability of oocyte donors offers a good model for the evaluation of the endometrium because their hormonal milieu after ovarian stimulation approximates closely that of infertile women undergoing controlled ovarian hyperstimulation (COH) for in vitro fertilization (IVF).
| Subjects and Methods |
|---|
|
|
|---|
Oocyte donors
Women from 2129 yr of age were eligible as oocyte donors. The selection process included an extensive questionnaire and psychological evaluation of the potential donors followed by a detailed physical examination and consultation about the process of oocyte donation by one of the physicians in the group. The risks of the procedure were discussed in detail, and written informed consents were obtained. All donors were screened for sexually transmitted diseases as well as for genetic conditions such as cystic fibrosis in accordance with the recommendations of the American Society for Reproductive Medicine (16). Women with a body mass index exceeding 28 kg/m2, history of pelvic inflammatory disease, sexually transmitted diseases, reproductive tract pathology, or other systemic diseases were excluded. At the time of their initial visit, they received a detailed explanation of the study protocol with particular emphasis on the risks associated with the endometrial biopsy and the use of steroids during their luteal phase. A written informed consent was obtained at that time. From August 1, 2003, to October 30, 2005, 20 healthy oocyte donors were initially recruited, 17 of which completed the study.
Stimulation protocol
Oocyte donors were stimulated with a GnRH antagonist protocol. Briefly, all donors had a baseline measurement of FSH and estradiol (E2) serum concentrations on the second day of their menstrual cycles after the discontinuation of oral contraceptive pills. In addition, a transvaginal sonogram was performed to rule out early follicular development and any anatomic anomalies. Providing that serum FSH was less than 10 mIU/ml and E2 was less than 60 pg/ml, ovarian stimulation was initiated with 225 IU recombinant FSH (Follitropin Alfa, Gonal-F; Serono Laboratories, Norwell, MA). A daily evening dose of ganirelix acetate (Antagon; Organon, West Orange, CA), 0.25 mg sc, was started either 6 d after the initiation of gonadotropins or at the time of identification of a leading follicle with mean diameter more than 13 mm and continued through the day of human chorionic gonadotropin (hCG). Thereafter, the dose of gonadotropins was adjusted in a step-down fashion according to follicular development by serial transvaginal ultrasound and serum E2 response. When at least three follicles reached a mean diameter of 18 mm, ovulation was triggered with a single im dose of 10,000 IU hCG (Profasi; Serono) or 20 U of a GnRH agonist administered in two doses 24 h apart. Transvaginal oocyte retrieval was performed under iv sedation 3436 h after hCG or the initial dose of GnRH agonist.
Randomization
Using a computer-generated model, study participants were randomized at the time of the retrieval into three groups: group I did not receive any luteal-phase support; group II received micronized progesterone in the form of vaginal suppositories, 200 mg every 6 h starting from the day after retrieval; and group III received a daily oral dose of 2 mg 17ß-estradiol in addition to the micronized progesterone.
Endometrial biopsies were performed with the use of a Pipelle catheter (Unimar, Wilton, CT) on the day of oocyte retrieval (d 14 of the ideal cycle) and then 3, 5, and 10 d later corresponding to ideal cycle d 17, 19, and 24. At least three endometrial biopsies were obtained from each donor and stored in liquid nitrogen. The specimens were then fixed in 10% formalin and subsequently embedded in paraffin for tissue microarray sectioning.
Serum levels of E2 and progesterone were measured on the day of retrieval and 3, 5, and 10 d after retrieval.
Tissue microarrays (TMA)
In this study, TMA were assembled from 61 paraffin-embedded endometrial samples. Three representative punches (each at 1.5 mm in diameter) were obtained from each specimen. The arrays encompass 183 tissue cores derived from 17 donors. All tissue cores were sectioned at 5 µm thickness and affixed to the TMA slides.
Immunohistochemistry
The expression of L-selectin ligands was examined by immunolocalization using an antibody (MECA-79) that binds to sulfated oligosaccharide epitopes of L-selectin ligands (14, 17). Briefly, the sections were dewaxed through descending grades of ethanol to distilled water and pretreated with Citra Buffer (Vector H3300; Vector Laboratories, Burlingame, CA) in a steamer (Black & Decker HA900, Hampstead, MD) at 90 C for 20 min. The tissue sections were then labeled with a rat antihuman L-selectin ligand monoclonal antibody (MECA-79; BD PharMingen Inc., San Diego, CA) at a concentration of 3.3 µg/ml with dilution of 1:30 in PBS. The sections were then incubated with a biotin-conjugated secondary antibody (goat antimouse Ig), which cross-reacts with the rat primary antibody at a dilution of 1:800 with PBS. Positive immunostaining was detected through interaction of avidin-biotin peroxidase (ABC) complex using a Ventana DAB Detection Kit (Ventana-Biotek Solutions Inc., Tucson, AZ). Isotype-specific irrelevant monoclonal antibody, generated against the human microphthalmia transcription factor (MiTF), was used as a primary antibody for the negative controls (18). Slides were subsequently counterstained with hematoxylin.
The intensity of staining in glandular and luminal epithelium of the tissue sections was assessed using the histological score (HSCORE) as described by others (19, 20). The HSCORE was calculated using the following equation: HSCORE =
Pi (i + 1), where i is the intensity of staining (1 = weak, 2 = moderate, and 3 = strong) and Pi is the percentage of stained epithelial cells for each intensity (0100%). The TMA tissue sections were scored by two independent investigators in a blinded fashion without the knowledge of sample identifiers, using a light microscope (Olympus, CH-2; Hitech Instruments, Inc., Edgemont, PA), and the average HSCORE was documented. This semiquantitative analysis has been shown to have a low intra- and interobserver variation (20).
Serum hormone assays
Blood samples were collected between 0700 and 0800 h. E2, FSH, LH, and progesterone were quantitatively measured with a solid-phase, two-site chemiluminescent enzyme immunometric assay with the Immulite Automated Analyzer (Diagnostic Products Corp., Los Angeles, CA). The intraassay statistics were calculated for samples from the results of 20 replicates in a single run. Interassay statistics were calculated for samples assayed in 20 different runs. Intra- and interassay coefficients of variation were 6.315.0 and 6.416.0%, respectively, for serum E2; 5.47.7 and 6.58.1%, respectively, for serum FSH; and 5.016.0 and 5.816.0%, respectively, for serum progesterone. The detection limit of the assay was approximately 0.2 ng/ml.
All specimens obtained were coded with a combination of letters and numbers that was used to identify the group, the individual donor within the group, and the order of the endometrial biopsy of the specific donor. The evaluation of the endometrial samples was performed by individuals who were blinded to the type of luteal-phase support and the sample order.
Statistical analysis
Statistical analysis was performed with a commercial statistical package SPSS version 10.0 (SPSS, Chicago, IL). One-way ANOVA was used for comparison of means between groups for normally distributed variables and Kruskal-Wallis when the assumption of normality was not applicable. A P value of <0.05 was considered significant.
| Results |
|---|
|
|
|---|
|
Serum progesterone levels were well above the ovulatory range of 10 ng/ml on the day of retrieval as expected in all groups regardless of the type of support (Table 2
). In groups II and III, progesterone levels remained elevated for up to 10 d after retrieval, whereas in group I, there was a precipitous drop by d 10. As a result, there was a significant decrease in the mean progesterone level in group I compared with the other two groups (P = 0.021). In contrast, E2 levels remained elevated in all groups up to d 5 after retrieval, and there was a gradual drop by d 10 regardless of the type of supplementation (Table 2
).
|
|
|
| Discussion |
|---|
|
|
|---|
Edwards et al. (21) were the first to postulate that inadequacy of the luteal phase after ovarian stimulation may be one of the main reasons for failure in IVF cycles. Subsequently, hormonal support of the luteal phase in patients undergoing COH became a common practice. Use of GnRH agonists in ovarian stimulation protocols to suppress premature LH surges increased the number of mature oocytes and embryos available for transfer and subsequently improved pregnancy rates (22, 23).
GnRH agonist use, however, has been associated with several undesired effects. Premature luteolysis and inadequate endometrial priming because of oversuppression of pituitary function are some of the unwanted consequences (24). A rapid increase in progesterone and E2 levels has been described during the early luteal phase of stimulation protocols using GnRH agonists. This is followed by premature luteolysis during the midluteal phase (25). Luteal-phase support with progesterone or hCG has been found to improve pregnancy rates when human menopausal gonadotropins were used in conjunction with GnRH agonists for ovarian stimulation and IVF (26).
A newer group of agents derived from additional substitutions on the original GnRH molecule has become clinically available (27, 28). These agents (GnRH antagonists) bind to GnRH receptors and, by competitive blockage, cause an immediate inhibition of gonadotropin release without the initial stimulatory response (29). Initial studies with GnRH antagonists showed suppression of serum LH levels within hours of administration followed by a rapid recovery of the pituitary function within 24 h after discontinuation (30, 31). Preliminary experience with ovarian stimulation protocols showed that these agents effectively prevent premature luteinization without compromising implantation and pregnancy rates (32, 33). Interestingly, in all the initial studies involving GnRH antagonists, luteal-phase support was routinely either in the form of progesterone or hCG, despite limited information about the quality of the luteal phase. The necessity of luteal-phase support after GnRH antagonist has been addressed in a recent prospective randomized study designed to evaluate luteal-phase characteristics in women who had final induction of oocyte maturation with recombinant hCG, recombinant LH, or Triptorelin in two respected centers. The study was interrupted prematurely because of an unacceptably low pregnancy and implantation rate in the group that had no luteal-phase support (34).
Thus far, studies comparing different luteal support protocols in women undergoing ovarian stimulation used implantation and pregnancy rates as endpoints (35, 36, 37). As expected, several confounding factors (such as oocyte quality and transfer technique as well as other preimplantation events) that could not be controlled for could have affected implantation and pregnancy rates independently of the endometrial quality. There is limited knowledge regarding the characteristics of the endometrium in COH protocols involving GnRH antagonists.
This study was designed to investigate possible alterations in the luteal-phase endometrium by using a surrogate marker of endometrial receptivity that has been found to play an important role in human implantation. To allow for a high-throughput tissue analysis, we have used a tissue chip approach (38) on paraffin-embedded samples instead of the traditional methods of immunostaining that are time consuming and rapidly exhausting of precious tissue resources.
L-selectin ligands were identified at the time of retrieval as well as 3 and 5 d after retrieval in all groups regardless of the type or lack of luteal-phase support. On d 10, however, there was a significant decrease in the presence of L-selectin ligands in all groups except in group III, which had received a combination of E2 and micronized progesterone. Administration of micronized progesterone alone did not seem to have any additional effect on the presence of L-selectin ligands compared with the group that had no support at all. The clinical relevance of these findings is not clearly understood. According to Beckers et al. (34), progesterone supplementation is necessary after ovarian stimulation with an antagonist protocol. From our data, there was no difference in the presence of L-selectin ligand around the time of implantation (d 5) between the group that received micronized progesterone and the one that had no support at all. In addition, at the same time, the mean serum levels of progesterone were similar in all groups (26.2 vs. 22.1 vs. 31.4 ng/ml, P = 0.639). It seems, therefore, unlikely that a potential beneficial effect of progesterone is exerted around that time.
Only on d 10 were progesterone levels in the nonsupplemented group severely decreased compared with the other two groups (P = 0.021), and at the same time there was a complete absence of L-selectin ligand staining in some of the samples. These findings indicate that maintenance of progesterone levels may be important for the events that follow the initial attachment of the embryo. Interestingly, on d 10 after retrieval, we observed a significant increase in the presence of L-selectin ligands only in the group that had received a combination of E2 and progesterone (group III). The fact that the combination of estrogen and progesterone was associated with persistence of L-selectin ligands on d 10 is interesting and should be evaluated further. The clinical significance of this finding and whether this may impact the events of early implantation is unclear. One could speculate that the addition of E2 may act either directly on the endometrium through estrogen receptors or indirectly by influencing the induction of endometrial progesterone receptors and augmenting the action of progesterone. Whether this contributes to an improvement in the endometrial receptivity and subsequent implantation and pregnancy rates is unclear. In stimulation protocols with GnRH agonists, there are data to support an improvement in implantation and pregnancy rates in the groups that had received a combination of estrogens and progesterone for luteal-phase support (39), but this is not a common practice by all.
In a previous publication (15) looking at L-selectin ligand expression in the natural cycle, we were able to show that there was a significant increase in both luminal and glandular endometrium during the early, mid, and late luteal phase compared with the follicular phase. In contrast, there was no difference in the expression of L-selectin ligand in either the glandular or the luminal endometrium throughout the luteal phase. In a subsequent publication (40) comparing the endometrium of donors supplemented with progesterone only, we have demonstrated a significant decrease in L-selectin expression during the mid to late luteal phase compared with unstimulated controls. Our findings in the current report support the notion that the combination of E2 and progesterone for luteal-phase support after ovarian stimulation with GnRH antagonists compares favorably to the unstimulated cycle in terms of L-selectin ligand expression. Evidence indicates that the attachment of an embryo to the endometrium depends upon the binding of L-selectins expressed by the trophoblast to L-selectin ligands expressed in the endometrium (9). The intensity of immunostaining for MECA-79 was increased in the luminal epithelium compared with the glandular epithelium in natural cycles (15). These results may be explained by the fact that the luminal epithelium serves as the initial contact point between the blastocyst and the endometrium, whereas glandular epithelium may be involved during placentation.
To the best of our knowledge, this is the first prospective randomized study to provide interesting information regarding the preparation of the endometrium in assisted reproduction cycles. There are certain limitations that need to be presented. The small number of oocyte donors could potentially limit the power of the study. Furthermore, the administration of a GnRH agonist in place of hCG for final oocyte maturation in cases at risk for ovarian hyperstimulation syndrome may impact the results. In addition, regarding the morphological evaluation of the endometrial samples, one has to take into account that we have used a semiquantitative method to evaluate the intensity, which can allow a certain degree of subjectivity. Finally, the compliance of the participants could represent yet another limiting factor because there were occasions in which some donors could not tolerate all four endometrial biopsies.
In summary, there is evidence that some type of luteal-phase support is necessary in ovarian stimulation protocols that include GnRH antagonists. Whether progesterone alone is adequate or a combination of progesterone and E2 is preferable is unclear. It seems that a combination of the two agents is associated with increased expression of L-selectin ligands. Larger prospective studies in the future may improve our understanding of endometrial physiology and provide more information about the best possible method for endometrial preparation after assisted reproductive technologies.
| Footnotes |
|---|
Disclosure statement: The authors have nothing to disclose.
First Published Online July 25, 2006
Abbreviations: COH, Controlled ovarian hyperstimulation; E2, estradiol; hCG, human chorionic gonadotropin; HEV, high endothelial venules; HSCORE, histological score; IVF, in vitro fertilization; MiTF, microphthalmia transcription factor; TMA, tissue microarrays.
Received March 7, 2006.
Accepted July 10, 2006.
| References |
|---|
|
|
|---|
1-globulin, an endometrial insulin-like growth factor (IGF) binding protein: evidence for two small molecular weight IGF binding proteins. Endocrinology 123:12021204[Abstract]
6 and ß4 integrin subunits throughout the menstrual cycle: no correlation with uterine receptivity. Fertil Steril 72:522526[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |