- Email: [email protected]
Endometrial hormone receptors and proliferation index in the periovulatory phase of stimulated embryo transfer cycles in comparison with natural cycles and relation to clinical pregnancy outcome
FERTILITY AND STERILITY威 VOL. 78, NO. 2, AUGUST 2002 Copyright ©2002 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.
THE ENDOMETRIUM
Endometrial hormone receptors and proliferation index in the periovulatory phase of stimulated embryo transfer cycles in comparison with natural cycles and relation to clinical pregnancy outcome Claire Bourgain, M.D.,a Filippo Ubaldi, M.D.,c Asimina Tavaniotou, M.D.,b Johan Smitz, Ph.D.,b Andre´ C. Van Steirteghem, Ph.D.,b and Paul Devroey, Ph.D.b University Hospital, Dutch-Speaking Brussels Free University (Vrije Universiteit Brussel), Brussels, Belgium
Received August 9, 2001; revised and accepted March 5, 2002. Supported by grant G.0060.00 of the FWO (Fund for Scientific Research Flanders– Belgium). Reprint requests: Claire Bourgain, M.D., Department of Pathology, University Hospital, Dutch-Speaking Brussels Free University (Vrije Universiteit Brussel), Laarbeeklaan 101, B-1090 Brussels, Belgium (FAX: 0032-2-4775085; E-mail: [email protected]). a Department of Pathology. b Centre for Reproductive Medicine. c Department of Reproductive Medicine, European Hospital, Rome, Italy. 0015-0282/02/$22.00 PII S0015-0282(02)03228-4
Objective: To investigate the endometrial steroid receptors and proliferation index in GnRH analogue/hMGstimulated cycles in comparison with natural cycles and their relation to clinical pregnancy outcome. Design: Prospective observational study. Setting: Tertiary referral center. Patient(s): Twenty-seven stimulated patients with GnRH agonist and hMG. Twenty normo-ovulatory patients were the natural cycle controls. Intervention(s): Endometrial aspiration biopsies: in stimulated cycles on the day of oocyte retrieval within the ET cycle (Day OPU) (n ⫽ 20) or 2 days later (Day OPU ⫹ 2) (n ⫽ 7); in natural cycles on the natural day of ovulation (Day NO) (n ⫽ 10) or on the day of ovulation ⫹ 2 (Day NO ⫹ 2) (n ⫽ 10). Main Outcome Measure(s): Comparison of endometrial maturation, estrogen (ER) and P receptor (PR), and proliferation index by immunohistochemistry in natural and stimulated cycles, correlation with pregnancy outcome in stimulated cycles. Result(s): Stimulated cycles Day OPU showed significantly advanced endometrial maturation compared to natural cycles Day NO; stromal ER and glandular and stromal PR staining was lower in stimulated than in natural cycles, but higher on Day OPU than on Day NO ⫹ 2; proliferation index was lower in all stimulated cycles. Steroid receptors and proliferation index in stimulated cycles were unrelated to clinical pregnancy occurrence. Conclusion(s): Compared to natural cycles, ovarian stimulation induced an imbalance in endometrial ER and PR and led to a profound antimitotic effect in the peri-ovulatory phase. These parameters were, however, not predictive of clinical pregnancy in cycles with ET. (Fertil Steril威 2002;78:237– 44. ©2002 by American Society for Reproductive Medicine.) Key Words: Endometrium, ovarian stimulation, steroid receptors, proliferation index, transfer cycle
After ovarian stimulation for IVF, relatively low implantation rates have been reported despite improved embryo culture conditions and luteal phase supplementation. Several alterations in endometrial maturation have been found in these types of cycles, and both glandular advancement and retardation have been observed using morphological and morphometrical criteria (1, 2). Supraphysiological steroid serum concentrations in the early pre- and postovulatory phases have been postulated to be responsible for an adverse effect on endometrial develop-
ment and embryo implantation. However, it remains unclear which endometrial alterations may affect receptivity. An altered expression of steroid hormone receptors in the endometrium has been found in cycles stimulated with GnRH agonists (GnRH-a) and hMG. Low estrogen receptor (ER) content and a variability of P receptor (PR) content have been described depending on the P exposure (3). In turn, these alterations may affect steroid-related endometrial cell proliferation and differentiation. After binding to the nuclear receptors, steroid receptors activate their transcription genes. 237
Estrogen (E) has been shown to be mitogenic in inducing a rapid onset of DNA synthesis in both glands and stroma, whereas P modulates E action in the glands by down-regulating both ER and PR content. In the stroma, the mitogenic effect of P is mediated by up-regulating both PR-A and PR-B isoforms of the receptor (4, 5). Endometrial cell division can be assessed by immunostaining the proliferating cells. In a previous study (6), using light microscopical evaluation, we demonstrated secretory transformation of the endometrium on the day of ovum pick-up (OPU) in GnRH-a/ hMG-stimulated cycles with ET. Endometrial dating could not be predicted from the absolute value of serum P concentration. The establishment of a clinical pregnancy after ET in these cycles was unrelated to P serum concentrations, but correlated significantly with endometrial dating, as no pregnancy was observed if the endometrial dating was advanced for more than 3 days. At present, no reports are available on the endometrial steroid receptors and their mitogenic effect in relation to pregnancy outcome in stimulated cycles. The aim of the present study was, first, to compare the endometrial steroid receptor immunolocation and proliferation index in GnRH-a/hMG-stimulated cycles to those in natural cycle controls and second, to correlate these parameters with the occurrence of a clinical pregnancy in the stimulated cycles where ET was performed 2 days after the endometrial biopsy.
MATERIALS AND METHODS Patients All patients were between 24 and 38 years old (mean age, 31.1 years), with basal FSH ⬍10 IU/L, a normal cycle length, no known female infertility factor, and no history of failed IVF attempts. Ovarian stimulation was performed in 27 patients for intracytoplasmic sperm injection (ISCI) due to male infertility. Of those patients, 20 underwent ET. The GnRH-a Buserelin and hMG was used in a long protocol as described previously (6, 7). Cycle monitoring, oocyte retrieval, oocyte and semen preparation, ICSI procedure, embryo classification, ET procedure, and luteal phase supplementation have all been described in detail in other publications (6, 8, 9). Resulting pregnancies were monitored by serum hCG (Hybritech, Sart Tilman, Belgium) and ultrasound evaluation of the uterine cavity. A clinical pregnancy was defined as an intrauterine gestational sac with embryonic heart activity visible at 7 weeks by ultrasonography. Twenty control patients were monitored throughout a natural cycle with daily serum blood samples for E2, P, FSH, and LH from the eighth day of the menstrual cycle onward. When E2 values were more than 150 ng/L, an ultrasound 238
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
scan was done to detect a growing follicle. Daily blood samples were taken until the detection of the peak value of the LH surge. The LH surge was defined as a serum value of LH exceeding 14 IU/L. The LH values were evaluated in addition to E2 and P to determine the peak value of the LH surge: a plateau or decrease in E2 values and an increase in serum P were considered as confirmatory variables. The day after the LH peak was considered as the day of ovulation (10, 11). The daily blood samples were taken until the third day after the LH peak and every 3 days for the rest of the luteal phase, until menstruation. Serum steroids were measured by commercially available radioimmunoassays and gonadotrophins were measured using specific monoclonal immunoradiometric assays for FSH and LH and expressed in international units per liter (Second International Standard Preparation for LH 80/552 and Second International Reference Preparation for FSH 78/549). Steroid serum concentrations were expressed in nanograms per liter for E2 and in micrograms per liter for P (conversion factor to SI unit 3.671 for E2 and 3.180 for P).
Endometrial Biopsies All biopsy specimens were obtained with an aspiration biopsy device (Pipelle de Cornier, Prodimed, Neuilly-enThelle, France or Z-sampler, ZSI, Chatsworth, CA). In stimulated cycles, biopsies were taken on the day of oocyte retrieval (Day OPU) (n ⫽ 20) or 2 days after oocyte retrieval (Day OPU ⫹ 2) (n ⫽ 7). In control natural cycles, biopsies were taken either on the estimated day of natural ovulation (Day NO) (n ⫽ 10) (determined as the day after the LH peak) or 2 days after ovulation (Day NO ⫹ 2) (n ⫽ 10) (determined as the third day after the LH peak) (10, 11). The biopsy specimens were fixed in 10% neutral buffered formalin, embedded in paraffin, and cut into 3-m-thick sections for light microscopical evaluation or immunohistochemical analysis (1). For endometrial analysis, observers were blind for cycle type and IVF outcome. Endometrial dating was performed according to the criteria of Noyes et al. (12) by two independent observers. Dating was recorded as days from the day of ovulation. The endometrium was considered as histologically in phase when the histological dating corresponded to the expected day ⫾1 day. The day of ovulation was considered day 0 of the luteal phase in natural cycles. In stimulated cycles, the day of oocyte retrieval was considered equivalent to the ovulation day and thus also as day 0 of the luteal phase. Immunohistochemistry was performed using commercially available antibodies for ␣-ER, both PR isoforms A and B, and cell proliferation (Ki67) (Clone 1D5 at a dilution of 1:50, Clone 1A6 at a dilution of 1:200, and Clone MIB1 at a dilution of 1:100, respectively, from Immunotech, Marseille, France), with the labeled streptavidin-biotin method. Vol. 78, No. 2, August 2002
FIGURE 1 Details of endometrial histological dating (Y-axis) of natural cycles (A) and stimulated cycles (B) expressed in function of the expected chronological luteal phase day (X-axis). Each patient is represented by the value of a small circle, the size of the circles representing all patients with identical dating results. The label in the circle is number of patients with identical dating results.
FIGURE 2 Immunohistochemistry for ␣-E (Mab 1D5) and P receptors (Mab 1A6), stimulated cycle, day of OPU. (a), Estrogen receptors: the endometrial glandular cells show a weak to moderate nuclear staining (arrow). The stromal cells are negative. (b), Progesterone receptors: the glandular cells are intensively stained (arrow). The stromal staining is moderate (arrowhead). The scale bar represents 50 m.
Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
Deparaffinated, rehydrated sections were incubated with 3% H2O2 in methanol for 30 minutes to block endogenous peroxidase activity. Antigen retrieval was done by microwave heating three times for 3 minutes in citrate buffer at pH 7.6. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
The sections were rinsed in phosphate-buffered saline, blocked with 10% normal goat serum for 30 minutes, and incubated with the primary antibodies diluted in phosphatebuffered saline overnight in a wet chamber at 4°C.
negative tissue sections and by substituting isotypic mouse serum for the primary antibody.
A second biotinylated goat anti-mouse antibody at a dilution of 1:300 and streptavidin/peroxidase conjugate were used according to the instructions of the manufacturer (Amersham Pharmacia Biotec, Buckinghamshire, UK). A chromogenic precipitate was obtained by immersion in diaminobenzidine solution for 7 minutes and in 0.5% CuSO4 for 5 minutes. Sections were counterstained with hematoxylin before mounting.
The immunohistochemical signal was reported by analyzing the functional endometrial layer in all 47 biopsies. The staining was recorded in epithelial cells from the endometrial glands (excluding surface epithelium) and in stromal cells (excluding vessel wall). In each of the 47 endometrial biopsy samples and for each of the three monoclonal antibodies, 1,000 glandular cell nuclei and 1,000 stromal cell nuclei were observed.
Positive controls were tissue sections containing the relevant antigens. Negative controls were obtained from known
For hormone receptors, the staining was assessed according to the scoring system reported previously (1). The stain-
FERTILITY & STERILITY威
239
TABLE 1 Area under the curve of serum P concentrations (mean and SD) in natural and stimulated cycles. Natural cycles Day NO (n ⫽ 10) 2.4 (0.78)
Stimulated cycles
Day NO ⫹ 2 (n ⫽ 10) 13.1 (4.97)
Day OPU (n ⫽ 20) 14.2 (4.27)
P value
Day OPU ⫹ 2 (n ⫽ 7) 79.4 (24.1)
⬍.001
P value represents the difference between natural and stimulated cycle groups with a two-way analysis of variance. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
ing was scored per nucleus from 0 (negative) to 3 (strongly positive). The mean score of 1,000 glandular and 1,000 stromal nuclei was recorded as the staining intensity of the glands and the stroma per endometrial sample.
dating, 2.3; range, 0 –3). Biopsies performed 2 days later (OPU ⫹ 2) were histologically in phase (mean dating, 2.71; range, 2–3), but displayed rare glandular mitotic figures and onset of stromal edema.
For Ki67, the proliferation index was reported as the number of positive cells in 1,000 glandular and 1,000 stromal cells per sample. All nuclear staining was considered as positive.
In natural cycles, all biopsies taken were histologically in phase on both study days. Biopsies taken on Day NO (estimated day of ovulation) showed an end-proliferative histological image in 10 of 10 cycles (mean dating, 0; range, 0) as expected and biopsies taken on Day NO ⫹ 2 showed exact luteal phase day 2 maturation in 7 of 10 cycles (mean dating, 2.1; range, 2–3); none of the biopsies were advanced or delayed by more than 1 day.
Statistical Analysis Data were analyzed with an ANOVA two-way analysis of variance and Wilcoxon signed rank tests as appropriate. Statistical significance was defined as P⬍.05.
Ethical Approval This study protocol was approved by the Institutional Review Board of the Vrije Universiteit Brussel. All patients gave informed consent.
RESULTS Endometrial Dating In all, 47 endometrial biopsies were analyzed (27 from stimulated and 20 from natural cycles). According to Noyes’ criteria, the majority of endometria taken on the day of oocyte retrieval (Day OPU) in stimulated cycles showed an advanced maturation. Secretory changes consistent with histological features appearing in the second and third day of the luteal phase were observed in 18 of 20 cycles (mean
Endometrial maturation in stimulated cycles Day OPU was significantly more advanced than in natural cycle Day NO (P⬍.001), but similar to natural cycle Day NO ⫹ 2 (Figure 1).
Serum E2 and P Concentrations The serum E2 concentration was higher in stimulated cycles on the day of OPU compared to natural cycles on the day of ovulation and on both days before OPU and ovulation (P⬍.001). Serum P concentrations, expressed as area under the curve from 2 days before OPU or ovulation until the day of biopsy, were significantly higher in stimulated than in natural cycles (P⬍.001) (Table 1). In stimulated cycles, the serum P concentrations on the day of hCG injection (mean, 0.57 g/L [SD 0.08]) were similar to those for natural cycles on the day of LH surge (mean, 0.59 g/L [SD 0.09]).
TABLE 2 Staining intensity of endometrial estrogen and progesterone receptors. Natural cycles
ERg ERs PRg PRs
Stimulated cycles
Day NO (n ⫽ 10)
Day NO ⫹ 2 (n ⫽ 10)
Day OPU (n ⫽ 20)
Day OPU ⫹ 2 (n ⫽ 7)
P values
1.27 (1.1–1.59) 0.40 (0.28–0.68) 2.65 (2.4–2.76) 2.01 (1.9–2.12)
1.14 (0.92–1.25) 0.09 (0–0.12) 2.02 (1.78–2.2) 1.45 (1.24–1.69)
0.94 (0.39–1.48) 0.16 (0.08–0.3) 2.4 (2.06–2.67) 1.82 (1.11–2.04)
0.78 (0.6–0.9) 0.01 (0–0.1) 1.52 (1.2–2.1) 0.86 (0.63–1.2)
⬍.001 ⬍.001 ⬍.001 ⬍.001
Values are expressed as mean and range in glandular (g) and stromal (s) cells per patient group. P values express the difference per variable between natural and stimulated cycle groups with a two-way analysis of variance. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
240
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
Vol. 78, No. 2, August 2002
Immunohistochemistry ␣-Estrogen Receptor In stimulated cycles in both cycle days, a weak nuclear immunoreactivity was found for glandular endometrial ER. In the stromal cells, the signal ER was weak to absent (Figure 2a).
FIGURE 3 Ki67 (MabMib1) immunohistochemical staining of proliferating cells in glandular (arrows) and stromal (arrowhead) cells of natural cycles, day 2 (a) and stimulated cycles, day of OPU (b). The scale bar represents 100 m.
Stimulated endometria Day OPU showed a nonsignificant trend toward lower glandular ER than control endometria Day NO (P⫽.068) and Day NO ⫹ 2 (P⫽.08). Stromal ER were significantly lower than in controls Day NO (P⫽.001), but higher than in controls Day NO ⫹ 2 (P⫽.01). Stimulated endometria Day OPU ⫹ 2 contained significant lower glandular ER and stromal ER than controls on both days (P⬍.01 in all cases). The results are presented in Table 2.
Progesterone Receptors A and B In both stimulated and natural cycles, endometrial PR nuclear immunoreactivity was strong in the glands and moderate in the stroma (Figure 2b). In the stimulated endometria on Day OPU, the glandular PR and stromal PR content were significantly lower than in controls on Day NO (P⫽.002 in both cell types), but higher than in controls on Day NO ⫹ 2 (P⫽.001 and .003, respectively). In the stimulated endometria on Day OPU ⫹ 2, glandular PR and stromal PR staining intensity was much lower than in all control endometria (P⬍.01). The PR content for each group is detailed in Table 2.
Proliferation Index A nuclear immunoreactivity for the proliferation marker Ki67 was present in endometrial glandular and stromal cells (Figure 3). Proliferation in endometrial glands and stroma was significantly lower in both biopsy groups from stimulated cycles compared to natural cycle controls on Day NO and Day NO ⫹ 2 (P⬍.001 in all conditions) (Table 3).
Steroid Receptors, Proliferation Index, and Pregnancy Outcome After ET In the 20 stimulated patients who had an endometrial biopsy on the day of oocyte retrieval, ET was performed 2 days after the biopsy and seven clinical pregnancies ensued. Embryo quality as assessed by morphological parameters (9) was similar in pregnant and nonpregnant patients. In the cycles with ET, no difference in endometrial ER and PR immunostaining was detected on the day of oocyte retrieval in patients with (n ⫽ 7) or without (n ⫽ 13) clinical pregnancy. On this cycle day, the proliferation index in both glandular and stromal cell types was also unrelated to pregnancy outcome.
DISCUSSION In the present study, more advanced secretory endometrial maturation combined with reduced ER and PR and a low proliferation index in glands and stroma was found in FERTILITY & STERILITY威
Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
GnRH-a/hMG-stimulated cycles on the day of oocyte retrieval than in natural cycles (controls) on the day of ovulation. In stimulated cycles where an endometrial biopsy was taken 2 days after oocyte retrieval, a further reduction in steroid receptors and proliferation was observed, despite a similar histological maturation as compared to natural cycles on day 2 after ovulation. At present, endometrial maturation and secretory features in stimulated cycles have been assessed according to the criteria used for the natural cycle, with the day of oocyte retrieval being considered as equivalent to the natural day of ovulation. On the day of oocyte retrieval in donor-stimulated cycles, advanced endometrial development was reported by some 241
TABLE 3 Proliferation index in endometrial glandular and stromal cells. Natural cycles
Ki67g Ki67s
Stimulated cycles
Day NO (n ⫽ 10)
Day NO ⫹ 2 (n ⫽ 10)
Day OPU (n ⫽ 20)
Day OPU ⫹ 2 (n ⫽ 7)
P values
315 (49–430) 283.7 (32–420)
441 (152–738) 228 (80–343)
49.3 (3–170) 41.4 (1–135)
104.1 (11–372) 104.2 (71–178)
⬍.001 ⬍.001
Values are expressed as mean and range in glandular (g) and stromal (s) cells per patient group. P values refer to the difference for each variable between the natural and stimulated cycle groups with a two-way analysis of variance. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
investigators, which was at least partially related to high P serum concentrations on the day after hCG injection (13, 14). In other reports, in-phase as well as delayed endometrial dating was found (15), but these stimulated cycles were obtained from women with endometrial polyps where some endometrial resistance to steroids cannot be excluded. However, in our previous study, we have demonstrated that exact endometrial development on the day of OPU in stimulated cycles could not be predicted solely from the serum P concentrations, because the endometrial dating was not related to the absolute value of serum P increase (6). In the present study, by comparing endometrial dating in stimulated cycles with precisely timed natural cycles, we found a rapid onset of secretory changes in stimulated cycles on the day of OPU corresponding at least to day 2 after ovulation in natural cycles. This advanced maturation was no longer visible in stimulated cycles on the day OPU ⫹ 2. This suggests that premature secretory changes were induced by the rapid increase in serum P between the day of hCG injection and the day of oocyte retrieval, rather than by the number of days or total amount of P exposure. In stimulated cycles, PR have been reported to be reduced in the early luteal phase (3, 13, 16). This premature PR decrease was consistent with an early high P exposure. In our series, PR content in both glandular and stromal cells in stimulated cycles was significantly lower than in natural cycles, and correlated with the higher integrated serum P concentrations. However, the immunohistochemical signal in stimulated cycles on Day OPU was significantly higher than for a natural luteal phase day 2. This would mean that in terms of PR, the stimulated endometrium on Day OPU is less advanced than would be expected from the endometrial dating and is still able to respond to later serum P influence. In natural cycles, a decrease of glandular PR occurs within a short time interval in the mid-luteal phase. The presence of relatively high PR in the periovulatory phase of stimulated cycles may explain the effect of luteal supplementation in these cycles. Indeed, on day 7 of the luteal phase, a decrease in glandular PR was observed more fre242
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
quently in supplemented cycles than in nonsupplemented cycles (17). Reports on endometrial ER in stimulated cycles are less clear, as both overall decreases and glandular ER increases have been described (3, 15). Differences in the choice of the monoclonal antibody and types of staining interpretation may explain the different results. With immunohistochemistry, only a weak staining intensity for ER is found in the early luteal phase, and minimal changes may be missed due to signal interpretation errors. In the present study, the lower ER in stimulated cycles on the day of OPU than in natural cycle controls on the day of ovulation was mainly observed in the stroma, but failed to reach a significant difference in the glands. As for PR, the staining intensity of the stromal ER in the stimulated cycles was higher than that of luteal phase day 2 of a natural cycle. These findings suggest a relative imbalance in ER and PR content of the endometrium in stimulated cycles compared to their natural cycle counterparts. The present study was carried out with antibodies recognizing the ␣-ER and both A and B isoforms of the PR. Recently, a new isoform of ER, -ER, has been demonstrated to be present in the endometrium throughout the menstrual cycle, with a presumed function in endometrial stromal cell decidualization (18). Furthermore, it has been demonstrated that the isoforms of the PR carry out opposing activities (5). The function of the individual isoforms of steroid receptors in stimulated cycles remains to be investigated. One of the main actions of E2 in the endometrium is the induction of mitotic activity in glandular and stromal cells through ␣-ER action, as found in ␣-ER null mice (19). Comparing natural cycle endometrium to stimulated cycles in cross-over studies on day 4 and day 10 of the luteal phase (20), no difference in glandular proliferation index was found between the two cycles. In our data, the proliferation index was consistently lower in stimulated cycles than in natural controls. The periovulatory period corresponds to the highest mitotic activity in the Vol. 78, No. 2, August 2002
endometrium, which decreases rapidly in later cycle days (21). This may explain why differences were not noted in mid-luteal phase studies. In our results, the antimitotic effect of stimulation was by far higher than expected from the steroid receptors. Such a discrepancy between proliferation index and ER expression has been also described after pure antiestrogen drug use (22), suggesting that even small changes in the steroid receptor content may considerably affect cell proliferation. An early decrease in glandular proliferation could explain why the advanced maturation on Day OPU ⫹ 2 was no longer visible in stimulated cycles in our study. This phenomenon can also be responsible for the glandular delay observed in the mid-luteal phase of stimulated cycles (23). From the endometrial morphological and immunohistochemical data comparing natural and stimulated cycles in the present study, it might be hypothesized that ovarian stimulation leads to functional endometrial alterations in the postovulatory phase, especially affecting the proliferative potential of the endometrium. As stimulated and natural cycles are extremely difficult to match on a timeline, differences with respect to a particular cycle day are to be considered with caution. From our results, the day of oocyte retrieval might not be the best baseline to address the luteal phase onset in stimulated cycles. Further studies examining the endometrial changes in the days preceding oocyte retrieval are mandatory to clarify this issue. In the present study, we performed an endometrial biopsy in stimulated cycles with ET. From previous work in such cycles, we demonstrated that advanced endometrial dating of more than 3 days on the day of oocyte retrieval correlated negatively with the establishment of a clinical pregnancy after ET (6). Data recording endometrial evaluation in cycles with ET are exceedingly rare and reports on steroid receptors in this situation are lacking. Using ultrasonography, a predictive value of the endometrial echogenicity on the day of oocyte retrieval was reported with regard to implantation success. In natural cycles, this triple line echogenic pattern correlated with high endometrial stromal PR levels (23, 24). From studies comparing the endometrial morphology in natural cycles preceding the stimulated ones with ET, natural cycle endometria with a higher mitotic activity led more frequently to pregnancies in later IVF cycles (25). It has been suggested that early down-regulation of the endometrial steroid receptors in stimulated cycles coinciding with epithelial expression of pinopodes may advance the endometrial implantation window by an average of 2 days (13). In the present work we were not able to demonstrate any association between the endometrial immunohistochemical signal for ER, PR, and proliferation index and pregnancy FERTILITY & STERILITY威
outcome in stimulated cycles. This might be due to the limited number of cycles studied. From the differences observed in the present study between natural control cycles and stimulated cycles resulting in a pregnancy, it can also be hypothesized that the endometrium can accommodate a wide range in periovulatory steroid receptor content and proliferation without affecting later luteal phase receptivity. In conclusion, ovarian stimulation with GnRH-a/hMG induces precocious secretory endometrial transformation around the time of oocyte retrieval. Compared to natural cycles, there appears to be an imbalance between endometrial steroid receptor content, proliferation index, and maturation in the peri- and postovulatory phases of stimulated cycles. In stimulated cycles with ET, neither the endometrial steroid receptor content nor the proliferation index on the day of oocyte retrieval seems to be related to the establishment of clinical pregnancies.
Acknowledgments: The authors thank Nicole Buelens and Anne Nuyts, laboratory staff from the Department of Pathology, AZ-VUB Brussels, Belgium, for technical assistance, Nadia Verbruggen, statistician, Centre for Reproductive Medicine, AZ-VUB Brussels, Belgium, for advice with statistical analysis, Frank Winter from the Language Education Centre, VUB Brussels, Belgium, for language editing, and Sofie De Moor, secretariat, Centre for Reproductive Medicine, AZ-VUB Brussels, Belgium, for secretarial work.
References 1. Bourgain C, Smitz J, Camus M, Erard P, Devroey P, Van Steirteghem AC, et al. Human endometrial maturation is markedly improved after luteal supplementation of gonadotrophin-releasing hormone analogue/ human menopausal gonadotrophin stimulated cycles. Hum Reprod 1994;9:32– 40. 2. Meyer WR, Novotny DB, Fritz MA, Beyler SA, Wolf LJ, Lessey BA. Effect of exogenous gonadotropins on endometrial maturation in oocyte donors. Fertil Steril 1999;71:109 –14. 3. Hadi FH, Chantler E, Anderson E, Nicholson R, McClelland RA, Seif MW. Ovulation induction and endometrial steroid receptors. Hum Reprod 1994;9:2405–10. 4. Salmi A, Pakarinen P, Peltola AM, Rutanen EM. The effect of intrauterine levonorgestrel use on the expression of c-JUN, oestrogen receptors, progesterone receptors and Ki-67 in human endometrium. Mol Hum Reprod 1998;4:1110 –5. 5. Tseng L, Zhu HH. Regulation of progesterone receptor messenger ribonucleic acid by progestin in human endometrial stromal cells. Biol Reprod 1997;57:1360 – 6. 6. Ubaldi F, Bourgain C, Tournaye H, Smitz J, Van Steirteghem A, Devroey P. Endometrial evaluation by aspiration biopsy on the day of oocyte retrieval in the embryo transfer cycles in patients with serum progesterone rise during the follicular phase. Fertil Steril 1997;67: 521– 6. 7. Smitz J, Ron-El R, Tarlatzis BC. The use of gonadotrophin releasing hormone agonists for in vitro fertilization and other assisted procreation techniques: experience from three centers. Hum Reprod 1992;S1:49 – 66. 8. Van Steirteghem AC, Nagy Z, Joris H, Liu J, Staessen C, Smitz J, et al. High fertilization and implantation rates after intracytoplasmic sperm injection. Hum Reprod 1993;8:1061– 6. 9. Staessen C, Van den Abbeel E, Janssenswillen C, Devroey P, Van Steirteghem A. Controlled comparison of Earle’s balanced salt solution with Me´ ne´ zo B2 medium for in-vitro fertilization performance. Hum Reprod 1994;9:1915–9. 10. WHO task force. Temporal relationships between ovulation and defined changes in the concentration of plasma estradiol-17, luteinizing hor-
243
11.
12. 13.
14. 15.
16.
17. 18.
mone, follicle-stimulating hormone, and progesterone. Am J Obstet Gynecol 1980;138:383–90. Li TC, Rogers AW, Lenton EA, Dockery P, Cooke I. A comparison between two methods of chronological dating of human endometrial biopsies during the luteal phase, and their correlation with histologic dating. Fertil Steril 1987;48:928 –32. Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950;1:3–25. Develioglu OH, Hsiu JG, Nikas G, Toner JP, Oehninger S, Jones HW. Endometrial estrogen and progesterone receptor and pinopode expression in stimulated cycles of oocyte donors. Fertil Steril 1999;71:1040 –7. Kolb BA, Najmabadi S, Paulson RJ. Ultrastructural characteristics of the luteal phase endometrium in patients undergoing controlled ovarian hyperstimulation. Fertil Steril 1997;67:625–30. Lass A, Peat D, Avery S, Brinsden P. Histological evaluation of endometrium on the day of oocyte retrieval after gonadotrophin-releasing hormone agonist-follicle stimulating hormone ovulation induction for in-vitro fertilization. Hum Reprod 1998;13:3203–5. Noci I, Borri P, Coccia ME, Crisculoli L, Scarselli G, Messeri, G, et al. Hormonal patterns, steroid receptors and morphological pictures of endometrium in hyperstimulated IVF cycles. Eur J Obstet Gynecol Reprod Biol 1997;75:215–20. Seppala M, Tiitinen A. Endometrial responses to corpus luteum products in cycles with induced ovulation: theoretical and practical considerations. Hum Reprod 1995;S2:67–76. Lecce G, Meduri G, Ancelin M, Bergeron C, Perrot-Applanat M. Presence of estrogen receptor beta in the human endometrium through
244
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
19. 20.
21.
22.
23.
24.
25.
the cycle: expression in glandular, stromal, and vascular cells. J Clin Endocrinol Metab 2001;86:1379 – 86. Couse JF, Korach KS. Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 1999;20:358 – 417. Bebington C, Doherty FJ, Ndukwe G, Fleming SD. The progesterone receptor and ubiquitin are differentially regulated within the endometrial glands of the natural and stimulated cycle. Mol Hum Reprod 2000;6:264 – 8. Dowsett M, Howell R, Salter J, Thomas NM, Thomas EJ. Effects of the pure anti-oestrogen ICI 182780 on oestrogen receptors, progesterone receptors and Ki67 antigen in human endometrium in vivo. Hum Reprod 1995;10:262–7. Basir GS, O Ws, Yu Ng EH, Ho PC. Morphometric analysis of peri-implantation endometrium in patients having excessively high oestradiol concentrations after ovarian stimulation. Hum Reprod 2001; 16:435– 40. Sharara FI, Lim J, McClamrock HD. Endometrial pattern on the day of oocyte retrieval is more predictive of implantation success than the pattern or thickness on the day of hCG administration. J Assist Reprod Genet 1999;16:523– 8. Kim CH, Chae HD, Huh J, Kang BM, Chang YS, Nam JH. Relationship between endometrial estrogen and progesterone receptors, and sonographic endometrial appearance in the preovulatory phase. J Obstet Gynaecol Res 2000;26:95–101. Scemiczky G, Wramsby H, Johannisson E, Landgren BM. Importance of endometrial quality in women with tubal infertility during a natural menstrual cycle for the outcome of IVF treatment. J Assist Reprod Genet 1998;15:55– 61.
Vol. 78, No. 2, August 2002
THE ENDOMETRIUM
Endometrial hormone receptors and proliferation index in the periovulatory phase of stimulated embryo transfer cycles in comparison with natural cycles and relation to clinical pregnancy outcome Claire Bourgain, M.D.,a Filippo Ubaldi, M.D.,c Asimina Tavaniotou, M.D.,b Johan Smitz, Ph.D.,b Andre´ C. Van Steirteghem, Ph.D.,b and Paul Devroey, Ph.D.b University Hospital, Dutch-Speaking Brussels Free University (Vrije Universiteit Brussel), Brussels, Belgium
Received August 9, 2001; revised and accepted March 5, 2002. Supported by grant G.0060.00 of the FWO (Fund for Scientific Research Flanders– Belgium). Reprint requests: Claire Bourgain, M.D., Department of Pathology, University Hospital, Dutch-Speaking Brussels Free University (Vrije Universiteit Brussel), Laarbeeklaan 101, B-1090 Brussels, Belgium (FAX: 0032-2-4775085; E-mail: [email protected]). a Department of Pathology. b Centre for Reproductive Medicine. c Department of Reproductive Medicine, European Hospital, Rome, Italy. 0015-0282/02/$22.00 PII S0015-0282(02)03228-4
Objective: To investigate the endometrial steroid receptors and proliferation index in GnRH analogue/hMGstimulated cycles in comparison with natural cycles and their relation to clinical pregnancy outcome. Design: Prospective observational study. Setting: Tertiary referral center. Patient(s): Twenty-seven stimulated patients with GnRH agonist and hMG. Twenty normo-ovulatory patients were the natural cycle controls. Intervention(s): Endometrial aspiration biopsies: in stimulated cycles on the day of oocyte retrieval within the ET cycle (Day OPU) (n ⫽ 20) or 2 days later (Day OPU ⫹ 2) (n ⫽ 7); in natural cycles on the natural day of ovulation (Day NO) (n ⫽ 10) or on the day of ovulation ⫹ 2 (Day NO ⫹ 2) (n ⫽ 10). Main Outcome Measure(s): Comparison of endometrial maturation, estrogen (ER) and P receptor (PR), and proliferation index by immunohistochemistry in natural and stimulated cycles, correlation with pregnancy outcome in stimulated cycles. Result(s): Stimulated cycles Day OPU showed significantly advanced endometrial maturation compared to natural cycles Day NO; stromal ER and glandular and stromal PR staining was lower in stimulated than in natural cycles, but higher on Day OPU than on Day NO ⫹ 2; proliferation index was lower in all stimulated cycles. Steroid receptors and proliferation index in stimulated cycles were unrelated to clinical pregnancy occurrence. Conclusion(s): Compared to natural cycles, ovarian stimulation induced an imbalance in endometrial ER and PR and led to a profound antimitotic effect in the peri-ovulatory phase. These parameters were, however, not predictive of clinical pregnancy in cycles with ET. (Fertil Steril威 2002;78:237– 44. ©2002 by American Society for Reproductive Medicine.) Key Words: Endometrium, ovarian stimulation, steroid receptors, proliferation index, transfer cycle
After ovarian stimulation for IVF, relatively low implantation rates have been reported despite improved embryo culture conditions and luteal phase supplementation. Several alterations in endometrial maturation have been found in these types of cycles, and both glandular advancement and retardation have been observed using morphological and morphometrical criteria (1, 2). Supraphysiological steroid serum concentrations in the early pre- and postovulatory phases have been postulated to be responsible for an adverse effect on endometrial develop-
ment and embryo implantation. However, it remains unclear which endometrial alterations may affect receptivity. An altered expression of steroid hormone receptors in the endometrium has been found in cycles stimulated with GnRH agonists (GnRH-a) and hMG. Low estrogen receptor (ER) content and a variability of P receptor (PR) content have been described depending on the P exposure (3). In turn, these alterations may affect steroid-related endometrial cell proliferation and differentiation. After binding to the nuclear receptors, steroid receptors activate their transcription genes. 237
Estrogen (E) has been shown to be mitogenic in inducing a rapid onset of DNA synthesis in both glands and stroma, whereas P modulates E action in the glands by down-regulating both ER and PR content. In the stroma, the mitogenic effect of P is mediated by up-regulating both PR-A and PR-B isoforms of the receptor (4, 5). Endometrial cell division can be assessed by immunostaining the proliferating cells. In a previous study (6), using light microscopical evaluation, we demonstrated secretory transformation of the endometrium on the day of ovum pick-up (OPU) in GnRH-a/ hMG-stimulated cycles with ET. Endometrial dating could not be predicted from the absolute value of serum P concentration. The establishment of a clinical pregnancy after ET in these cycles was unrelated to P serum concentrations, but correlated significantly with endometrial dating, as no pregnancy was observed if the endometrial dating was advanced for more than 3 days. At present, no reports are available on the endometrial steroid receptors and their mitogenic effect in relation to pregnancy outcome in stimulated cycles. The aim of the present study was, first, to compare the endometrial steroid receptor immunolocation and proliferation index in GnRH-a/hMG-stimulated cycles to those in natural cycle controls and second, to correlate these parameters with the occurrence of a clinical pregnancy in the stimulated cycles where ET was performed 2 days after the endometrial biopsy.
MATERIALS AND METHODS Patients All patients were between 24 and 38 years old (mean age, 31.1 years), with basal FSH ⬍10 IU/L, a normal cycle length, no known female infertility factor, and no history of failed IVF attempts. Ovarian stimulation was performed in 27 patients for intracytoplasmic sperm injection (ISCI) due to male infertility. Of those patients, 20 underwent ET. The GnRH-a Buserelin and hMG was used in a long protocol as described previously (6, 7). Cycle monitoring, oocyte retrieval, oocyte and semen preparation, ICSI procedure, embryo classification, ET procedure, and luteal phase supplementation have all been described in detail in other publications (6, 8, 9). Resulting pregnancies were monitored by serum hCG (Hybritech, Sart Tilman, Belgium) and ultrasound evaluation of the uterine cavity. A clinical pregnancy was defined as an intrauterine gestational sac with embryonic heart activity visible at 7 weeks by ultrasonography. Twenty control patients were monitored throughout a natural cycle with daily serum blood samples for E2, P, FSH, and LH from the eighth day of the menstrual cycle onward. When E2 values were more than 150 ng/L, an ultrasound 238
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
scan was done to detect a growing follicle. Daily blood samples were taken until the detection of the peak value of the LH surge. The LH surge was defined as a serum value of LH exceeding 14 IU/L. The LH values were evaluated in addition to E2 and P to determine the peak value of the LH surge: a plateau or decrease in E2 values and an increase in serum P were considered as confirmatory variables. The day after the LH peak was considered as the day of ovulation (10, 11). The daily blood samples were taken until the third day after the LH peak and every 3 days for the rest of the luteal phase, until menstruation. Serum steroids were measured by commercially available radioimmunoassays and gonadotrophins were measured using specific monoclonal immunoradiometric assays for FSH and LH and expressed in international units per liter (Second International Standard Preparation for LH 80/552 and Second International Reference Preparation for FSH 78/549). Steroid serum concentrations were expressed in nanograms per liter for E2 and in micrograms per liter for P (conversion factor to SI unit 3.671 for E2 and 3.180 for P).
Endometrial Biopsies All biopsy specimens were obtained with an aspiration biopsy device (Pipelle de Cornier, Prodimed, Neuilly-enThelle, France or Z-sampler, ZSI, Chatsworth, CA). In stimulated cycles, biopsies were taken on the day of oocyte retrieval (Day OPU) (n ⫽ 20) or 2 days after oocyte retrieval (Day OPU ⫹ 2) (n ⫽ 7). In control natural cycles, biopsies were taken either on the estimated day of natural ovulation (Day NO) (n ⫽ 10) (determined as the day after the LH peak) or 2 days after ovulation (Day NO ⫹ 2) (n ⫽ 10) (determined as the third day after the LH peak) (10, 11). The biopsy specimens were fixed in 10% neutral buffered formalin, embedded in paraffin, and cut into 3-m-thick sections for light microscopical evaluation or immunohistochemical analysis (1). For endometrial analysis, observers were blind for cycle type and IVF outcome. Endometrial dating was performed according to the criteria of Noyes et al. (12) by two independent observers. Dating was recorded as days from the day of ovulation. The endometrium was considered as histologically in phase when the histological dating corresponded to the expected day ⫾1 day. The day of ovulation was considered day 0 of the luteal phase in natural cycles. In stimulated cycles, the day of oocyte retrieval was considered equivalent to the ovulation day and thus also as day 0 of the luteal phase. Immunohistochemistry was performed using commercially available antibodies for ␣-ER, both PR isoforms A and B, and cell proliferation (Ki67) (Clone 1D5 at a dilution of 1:50, Clone 1A6 at a dilution of 1:200, and Clone MIB1 at a dilution of 1:100, respectively, from Immunotech, Marseille, France), with the labeled streptavidin-biotin method. Vol. 78, No. 2, August 2002
FIGURE 1 Details of endometrial histological dating (Y-axis) of natural cycles (A) and stimulated cycles (B) expressed in function of the expected chronological luteal phase day (X-axis). Each patient is represented by the value of a small circle, the size of the circles representing all patients with identical dating results. The label in the circle is number of patients with identical dating results.
FIGURE 2 Immunohistochemistry for ␣-E (Mab 1D5) and P receptors (Mab 1A6), stimulated cycle, day of OPU. (a), Estrogen receptors: the endometrial glandular cells show a weak to moderate nuclear staining (arrow). The stromal cells are negative. (b), Progesterone receptors: the glandular cells are intensively stained (arrow). The stromal staining is moderate (arrowhead). The scale bar represents 50 m.
Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
Deparaffinated, rehydrated sections were incubated with 3% H2O2 in methanol for 30 minutes to block endogenous peroxidase activity. Antigen retrieval was done by microwave heating three times for 3 minutes in citrate buffer at pH 7.6. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
The sections were rinsed in phosphate-buffered saline, blocked with 10% normal goat serum for 30 minutes, and incubated with the primary antibodies diluted in phosphatebuffered saline overnight in a wet chamber at 4°C.
negative tissue sections and by substituting isotypic mouse serum for the primary antibody.
A second biotinylated goat anti-mouse antibody at a dilution of 1:300 and streptavidin/peroxidase conjugate were used according to the instructions of the manufacturer (Amersham Pharmacia Biotec, Buckinghamshire, UK). A chromogenic precipitate was obtained by immersion in diaminobenzidine solution for 7 minutes and in 0.5% CuSO4 for 5 minutes. Sections were counterstained with hematoxylin before mounting.
The immunohistochemical signal was reported by analyzing the functional endometrial layer in all 47 biopsies. The staining was recorded in epithelial cells from the endometrial glands (excluding surface epithelium) and in stromal cells (excluding vessel wall). In each of the 47 endometrial biopsy samples and for each of the three monoclonal antibodies, 1,000 glandular cell nuclei and 1,000 stromal cell nuclei were observed.
Positive controls were tissue sections containing the relevant antigens. Negative controls were obtained from known
For hormone receptors, the staining was assessed according to the scoring system reported previously (1). The stain-
FERTILITY & STERILITY威
239
TABLE 1 Area under the curve of serum P concentrations (mean and SD) in natural and stimulated cycles. Natural cycles Day NO (n ⫽ 10) 2.4 (0.78)
Stimulated cycles
Day NO ⫹ 2 (n ⫽ 10) 13.1 (4.97)
Day OPU (n ⫽ 20) 14.2 (4.27)
P value
Day OPU ⫹ 2 (n ⫽ 7) 79.4 (24.1)
⬍.001
P value represents the difference between natural and stimulated cycle groups with a two-way analysis of variance. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
ing was scored per nucleus from 0 (negative) to 3 (strongly positive). The mean score of 1,000 glandular and 1,000 stromal nuclei was recorded as the staining intensity of the glands and the stroma per endometrial sample.
dating, 2.3; range, 0 –3). Biopsies performed 2 days later (OPU ⫹ 2) were histologically in phase (mean dating, 2.71; range, 2–3), but displayed rare glandular mitotic figures and onset of stromal edema.
For Ki67, the proliferation index was reported as the number of positive cells in 1,000 glandular and 1,000 stromal cells per sample. All nuclear staining was considered as positive.
In natural cycles, all biopsies taken were histologically in phase on both study days. Biopsies taken on Day NO (estimated day of ovulation) showed an end-proliferative histological image in 10 of 10 cycles (mean dating, 0; range, 0) as expected and biopsies taken on Day NO ⫹ 2 showed exact luteal phase day 2 maturation in 7 of 10 cycles (mean dating, 2.1; range, 2–3); none of the biopsies were advanced or delayed by more than 1 day.
Statistical Analysis Data were analyzed with an ANOVA two-way analysis of variance and Wilcoxon signed rank tests as appropriate. Statistical significance was defined as P⬍.05.
Ethical Approval This study protocol was approved by the Institutional Review Board of the Vrije Universiteit Brussel. All patients gave informed consent.
RESULTS Endometrial Dating In all, 47 endometrial biopsies were analyzed (27 from stimulated and 20 from natural cycles). According to Noyes’ criteria, the majority of endometria taken on the day of oocyte retrieval (Day OPU) in stimulated cycles showed an advanced maturation. Secretory changes consistent with histological features appearing in the second and third day of the luteal phase were observed in 18 of 20 cycles (mean
Endometrial maturation in stimulated cycles Day OPU was significantly more advanced than in natural cycle Day NO (P⬍.001), but similar to natural cycle Day NO ⫹ 2 (Figure 1).
Serum E2 and P Concentrations The serum E2 concentration was higher in stimulated cycles on the day of OPU compared to natural cycles on the day of ovulation and on both days before OPU and ovulation (P⬍.001). Serum P concentrations, expressed as area under the curve from 2 days before OPU or ovulation until the day of biopsy, were significantly higher in stimulated than in natural cycles (P⬍.001) (Table 1). In stimulated cycles, the serum P concentrations on the day of hCG injection (mean, 0.57 g/L [SD 0.08]) were similar to those for natural cycles on the day of LH surge (mean, 0.59 g/L [SD 0.09]).
TABLE 2 Staining intensity of endometrial estrogen and progesterone receptors. Natural cycles
ERg ERs PRg PRs
Stimulated cycles
Day NO (n ⫽ 10)
Day NO ⫹ 2 (n ⫽ 10)
Day OPU (n ⫽ 20)
Day OPU ⫹ 2 (n ⫽ 7)
P values
1.27 (1.1–1.59) 0.40 (0.28–0.68) 2.65 (2.4–2.76) 2.01 (1.9–2.12)
1.14 (0.92–1.25) 0.09 (0–0.12) 2.02 (1.78–2.2) 1.45 (1.24–1.69)
0.94 (0.39–1.48) 0.16 (0.08–0.3) 2.4 (2.06–2.67) 1.82 (1.11–2.04)
0.78 (0.6–0.9) 0.01 (0–0.1) 1.52 (1.2–2.1) 0.86 (0.63–1.2)
⬍.001 ⬍.001 ⬍.001 ⬍.001
Values are expressed as mean and range in glandular (g) and stromal (s) cells per patient group. P values express the difference per variable between natural and stimulated cycle groups with a two-way analysis of variance. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
240
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
Vol. 78, No. 2, August 2002
Immunohistochemistry ␣-Estrogen Receptor In stimulated cycles in both cycle days, a weak nuclear immunoreactivity was found for glandular endometrial ER. In the stromal cells, the signal ER was weak to absent (Figure 2a).
FIGURE 3 Ki67 (MabMib1) immunohistochemical staining of proliferating cells in glandular (arrows) and stromal (arrowhead) cells of natural cycles, day 2 (a) and stimulated cycles, day of OPU (b). The scale bar represents 100 m.
Stimulated endometria Day OPU showed a nonsignificant trend toward lower glandular ER than control endometria Day NO (P⫽.068) and Day NO ⫹ 2 (P⫽.08). Stromal ER were significantly lower than in controls Day NO (P⫽.001), but higher than in controls Day NO ⫹ 2 (P⫽.01). Stimulated endometria Day OPU ⫹ 2 contained significant lower glandular ER and stromal ER than controls on both days (P⬍.01 in all cases). The results are presented in Table 2.
Progesterone Receptors A and B In both stimulated and natural cycles, endometrial PR nuclear immunoreactivity was strong in the glands and moderate in the stroma (Figure 2b). In the stimulated endometria on Day OPU, the glandular PR and stromal PR content were significantly lower than in controls on Day NO (P⫽.002 in both cell types), but higher than in controls on Day NO ⫹ 2 (P⫽.001 and .003, respectively). In the stimulated endometria on Day OPU ⫹ 2, glandular PR and stromal PR staining intensity was much lower than in all control endometria (P⬍.01). The PR content for each group is detailed in Table 2.
Proliferation Index A nuclear immunoreactivity for the proliferation marker Ki67 was present in endometrial glandular and stromal cells (Figure 3). Proliferation in endometrial glands and stroma was significantly lower in both biopsy groups from stimulated cycles compared to natural cycle controls on Day NO and Day NO ⫹ 2 (P⬍.001 in all conditions) (Table 3).
Steroid Receptors, Proliferation Index, and Pregnancy Outcome After ET In the 20 stimulated patients who had an endometrial biopsy on the day of oocyte retrieval, ET was performed 2 days after the biopsy and seven clinical pregnancies ensued. Embryo quality as assessed by morphological parameters (9) was similar in pregnant and nonpregnant patients. In the cycles with ET, no difference in endometrial ER and PR immunostaining was detected on the day of oocyte retrieval in patients with (n ⫽ 7) or without (n ⫽ 13) clinical pregnancy. On this cycle day, the proliferation index in both glandular and stromal cell types was also unrelated to pregnancy outcome.
DISCUSSION In the present study, more advanced secretory endometrial maturation combined with reduced ER and PR and a low proliferation index in glands and stroma was found in FERTILITY & STERILITY威
Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
GnRH-a/hMG-stimulated cycles on the day of oocyte retrieval than in natural cycles (controls) on the day of ovulation. In stimulated cycles where an endometrial biopsy was taken 2 days after oocyte retrieval, a further reduction in steroid receptors and proliferation was observed, despite a similar histological maturation as compared to natural cycles on day 2 after ovulation. At present, endometrial maturation and secretory features in stimulated cycles have been assessed according to the criteria used for the natural cycle, with the day of oocyte retrieval being considered as equivalent to the natural day of ovulation. On the day of oocyte retrieval in donor-stimulated cycles, advanced endometrial development was reported by some 241
TABLE 3 Proliferation index in endometrial glandular and stromal cells. Natural cycles
Ki67g Ki67s
Stimulated cycles
Day NO (n ⫽ 10)
Day NO ⫹ 2 (n ⫽ 10)
Day OPU (n ⫽ 20)
Day OPU ⫹ 2 (n ⫽ 7)
P values
315 (49–430) 283.7 (32–420)
441 (152–738) 228 (80–343)
49.3 (3–170) 41.4 (1–135)
104.1 (11–372) 104.2 (71–178)
⬍.001 ⬍.001
Values are expressed as mean and range in glandular (g) and stromal (s) cells per patient group. P values refer to the difference for each variable between the natural and stimulated cycle groups with a two-way analysis of variance. Bourgain. Steroid receptors and Ki67 in IVF cycles. Fertil Steril 2002.
investigators, which was at least partially related to high P serum concentrations on the day after hCG injection (13, 14). In other reports, in-phase as well as delayed endometrial dating was found (15), but these stimulated cycles were obtained from women with endometrial polyps where some endometrial resistance to steroids cannot be excluded. However, in our previous study, we have demonstrated that exact endometrial development on the day of OPU in stimulated cycles could not be predicted solely from the serum P concentrations, because the endometrial dating was not related to the absolute value of serum P increase (6). In the present study, by comparing endometrial dating in stimulated cycles with precisely timed natural cycles, we found a rapid onset of secretory changes in stimulated cycles on the day of OPU corresponding at least to day 2 after ovulation in natural cycles. This advanced maturation was no longer visible in stimulated cycles on the day OPU ⫹ 2. This suggests that premature secretory changes were induced by the rapid increase in serum P between the day of hCG injection and the day of oocyte retrieval, rather than by the number of days or total amount of P exposure. In stimulated cycles, PR have been reported to be reduced in the early luteal phase (3, 13, 16). This premature PR decrease was consistent with an early high P exposure. In our series, PR content in both glandular and stromal cells in stimulated cycles was significantly lower than in natural cycles, and correlated with the higher integrated serum P concentrations. However, the immunohistochemical signal in stimulated cycles on Day OPU was significantly higher than for a natural luteal phase day 2. This would mean that in terms of PR, the stimulated endometrium on Day OPU is less advanced than would be expected from the endometrial dating and is still able to respond to later serum P influence. In natural cycles, a decrease of glandular PR occurs within a short time interval in the mid-luteal phase. The presence of relatively high PR in the periovulatory phase of stimulated cycles may explain the effect of luteal supplementation in these cycles. Indeed, on day 7 of the luteal phase, a decrease in glandular PR was observed more fre242
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
quently in supplemented cycles than in nonsupplemented cycles (17). Reports on endometrial ER in stimulated cycles are less clear, as both overall decreases and glandular ER increases have been described (3, 15). Differences in the choice of the monoclonal antibody and types of staining interpretation may explain the different results. With immunohistochemistry, only a weak staining intensity for ER is found in the early luteal phase, and minimal changes may be missed due to signal interpretation errors. In the present study, the lower ER in stimulated cycles on the day of OPU than in natural cycle controls on the day of ovulation was mainly observed in the stroma, but failed to reach a significant difference in the glands. As for PR, the staining intensity of the stromal ER in the stimulated cycles was higher than that of luteal phase day 2 of a natural cycle. These findings suggest a relative imbalance in ER and PR content of the endometrium in stimulated cycles compared to their natural cycle counterparts. The present study was carried out with antibodies recognizing the ␣-ER and both A and B isoforms of the PR. Recently, a new isoform of ER, -ER, has been demonstrated to be present in the endometrium throughout the menstrual cycle, with a presumed function in endometrial stromal cell decidualization (18). Furthermore, it has been demonstrated that the isoforms of the PR carry out opposing activities (5). The function of the individual isoforms of steroid receptors in stimulated cycles remains to be investigated. One of the main actions of E2 in the endometrium is the induction of mitotic activity in glandular and stromal cells through ␣-ER action, as found in ␣-ER null mice (19). Comparing natural cycle endometrium to stimulated cycles in cross-over studies on day 4 and day 10 of the luteal phase (20), no difference in glandular proliferation index was found between the two cycles. In our data, the proliferation index was consistently lower in stimulated cycles than in natural controls. The periovulatory period corresponds to the highest mitotic activity in the Vol. 78, No. 2, August 2002
endometrium, which decreases rapidly in later cycle days (21). This may explain why differences were not noted in mid-luteal phase studies. In our results, the antimitotic effect of stimulation was by far higher than expected from the steroid receptors. Such a discrepancy between proliferation index and ER expression has been also described after pure antiestrogen drug use (22), suggesting that even small changes in the steroid receptor content may considerably affect cell proliferation. An early decrease in glandular proliferation could explain why the advanced maturation on Day OPU ⫹ 2 was no longer visible in stimulated cycles in our study. This phenomenon can also be responsible for the glandular delay observed in the mid-luteal phase of stimulated cycles (23). From the endometrial morphological and immunohistochemical data comparing natural and stimulated cycles in the present study, it might be hypothesized that ovarian stimulation leads to functional endometrial alterations in the postovulatory phase, especially affecting the proliferative potential of the endometrium. As stimulated and natural cycles are extremely difficult to match on a timeline, differences with respect to a particular cycle day are to be considered with caution. From our results, the day of oocyte retrieval might not be the best baseline to address the luteal phase onset in stimulated cycles. Further studies examining the endometrial changes in the days preceding oocyte retrieval are mandatory to clarify this issue. In the present study, we performed an endometrial biopsy in stimulated cycles with ET. From previous work in such cycles, we demonstrated that advanced endometrial dating of more than 3 days on the day of oocyte retrieval correlated negatively with the establishment of a clinical pregnancy after ET (6). Data recording endometrial evaluation in cycles with ET are exceedingly rare and reports on steroid receptors in this situation are lacking. Using ultrasonography, a predictive value of the endometrial echogenicity on the day of oocyte retrieval was reported with regard to implantation success. In natural cycles, this triple line echogenic pattern correlated with high endometrial stromal PR levels (23, 24). From studies comparing the endometrial morphology in natural cycles preceding the stimulated ones with ET, natural cycle endometria with a higher mitotic activity led more frequently to pregnancies in later IVF cycles (25). It has been suggested that early down-regulation of the endometrial steroid receptors in stimulated cycles coinciding with epithelial expression of pinopodes may advance the endometrial implantation window by an average of 2 days (13). In the present work we were not able to demonstrate any association between the endometrial immunohistochemical signal for ER, PR, and proliferation index and pregnancy FERTILITY & STERILITY威
outcome in stimulated cycles. This might be due to the limited number of cycles studied. From the differences observed in the present study between natural control cycles and stimulated cycles resulting in a pregnancy, it can also be hypothesized that the endometrium can accommodate a wide range in periovulatory steroid receptor content and proliferation without affecting later luteal phase receptivity. In conclusion, ovarian stimulation with GnRH-a/hMG induces precocious secretory endometrial transformation around the time of oocyte retrieval. Compared to natural cycles, there appears to be an imbalance between endometrial steroid receptor content, proliferation index, and maturation in the peri- and postovulatory phases of stimulated cycles. In stimulated cycles with ET, neither the endometrial steroid receptor content nor the proliferation index on the day of oocyte retrieval seems to be related to the establishment of clinical pregnancies.
Acknowledgments: The authors thank Nicole Buelens and Anne Nuyts, laboratory staff from the Department of Pathology, AZ-VUB Brussels, Belgium, for technical assistance, Nadia Verbruggen, statistician, Centre for Reproductive Medicine, AZ-VUB Brussels, Belgium, for advice with statistical analysis, Frank Winter from the Language Education Centre, VUB Brussels, Belgium, for language editing, and Sofie De Moor, secretariat, Centre for Reproductive Medicine, AZ-VUB Brussels, Belgium, for secretarial work.
References 1. Bourgain C, Smitz J, Camus M, Erard P, Devroey P, Van Steirteghem AC, et al. Human endometrial maturation is markedly improved after luteal supplementation of gonadotrophin-releasing hormone analogue/ human menopausal gonadotrophin stimulated cycles. Hum Reprod 1994;9:32– 40. 2. Meyer WR, Novotny DB, Fritz MA, Beyler SA, Wolf LJ, Lessey BA. Effect of exogenous gonadotropins on endometrial maturation in oocyte donors. Fertil Steril 1999;71:109 –14. 3. Hadi FH, Chantler E, Anderson E, Nicholson R, McClelland RA, Seif MW. Ovulation induction and endometrial steroid receptors. Hum Reprod 1994;9:2405–10. 4. Salmi A, Pakarinen P, Peltola AM, Rutanen EM. The effect of intrauterine levonorgestrel use on the expression of c-JUN, oestrogen receptors, progesterone receptors and Ki-67 in human endometrium. Mol Hum Reprod 1998;4:1110 –5. 5. Tseng L, Zhu HH. Regulation of progesterone receptor messenger ribonucleic acid by progestin in human endometrial stromal cells. Biol Reprod 1997;57:1360 – 6. 6. Ubaldi F, Bourgain C, Tournaye H, Smitz J, Van Steirteghem A, Devroey P. Endometrial evaluation by aspiration biopsy on the day of oocyte retrieval in the embryo transfer cycles in patients with serum progesterone rise during the follicular phase. Fertil Steril 1997;67: 521– 6. 7. Smitz J, Ron-El R, Tarlatzis BC. The use of gonadotrophin releasing hormone agonists for in vitro fertilization and other assisted procreation techniques: experience from three centers. Hum Reprod 1992;S1:49 – 66. 8. Van Steirteghem AC, Nagy Z, Joris H, Liu J, Staessen C, Smitz J, et al. High fertilization and implantation rates after intracytoplasmic sperm injection. Hum Reprod 1993;8:1061– 6. 9. Staessen C, Van den Abbeel E, Janssenswillen C, Devroey P, Van Steirteghem A. Controlled comparison of Earle’s balanced salt solution with Me´ ne´ zo B2 medium for in-vitro fertilization performance. Hum Reprod 1994;9:1915–9. 10. WHO task force. Temporal relationships between ovulation and defined changes in the concentration of plasma estradiol-17, luteinizing hor-
243
11.
12. 13.
14. 15.
16.
17. 18.
mone, follicle-stimulating hormone, and progesterone. Am J Obstet Gynecol 1980;138:383–90. Li TC, Rogers AW, Lenton EA, Dockery P, Cooke I. A comparison between two methods of chronological dating of human endometrial biopsies during the luteal phase, and their correlation with histologic dating. Fertil Steril 1987;48:928 –32. Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950;1:3–25. Develioglu OH, Hsiu JG, Nikas G, Toner JP, Oehninger S, Jones HW. Endometrial estrogen and progesterone receptor and pinopode expression in stimulated cycles of oocyte donors. Fertil Steril 1999;71:1040 –7. Kolb BA, Najmabadi S, Paulson RJ. Ultrastructural characteristics of the luteal phase endometrium in patients undergoing controlled ovarian hyperstimulation. Fertil Steril 1997;67:625–30. Lass A, Peat D, Avery S, Brinsden P. Histological evaluation of endometrium on the day of oocyte retrieval after gonadotrophin-releasing hormone agonist-follicle stimulating hormone ovulation induction for in-vitro fertilization. Hum Reprod 1998;13:3203–5. Noci I, Borri P, Coccia ME, Crisculoli L, Scarselli G, Messeri, G, et al. Hormonal patterns, steroid receptors and morphological pictures of endometrium in hyperstimulated IVF cycles. Eur J Obstet Gynecol Reprod Biol 1997;75:215–20. Seppala M, Tiitinen A. Endometrial responses to corpus luteum products in cycles with induced ovulation: theoretical and practical considerations. Hum Reprod 1995;S2:67–76. Lecce G, Meduri G, Ancelin M, Bergeron C, Perrot-Applanat M. Presence of estrogen receptor beta in the human endometrium through
244
Bourgain et al.
Steroid receptors and Ki67 in IVF cycles
19. 20.
21.
22.
23.
24.
25.
the cycle: expression in glandular, stromal, and vascular cells. J Clin Endocrinol Metab 2001;86:1379 – 86. Couse JF, Korach KS. Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 1999;20:358 – 417. Bebington C, Doherty FJ, Ndukwe G, Fleming SD. The progesterone receptor and ubiquitin are differentially regulated within the endometrial glands of the natural and stimulated cycle. Mol Hum Reprod 2000;6:264 – 8. Dowsett M, Howell R, Salter J, Thomas NM, Thomas EJ. Effects of the pure anti-oestrogen ICI 182780 on oestrogen receptors, progesterone receptors and Ki67 antigen in human endometrium in vivo. Hum Reprod 1995;10:262–7. Basir GS, O Ws, Yu Ng EH, Ho PC. Morphometric analysis of peri-implantation endometrium in patients having excessively high oestradiol concentrations after ovarian stimulation. Hum Reprod 2001; 16:435– 40. Sharara FI, Lim J, McClamrock HD. Endometrial pattern on the day of oocyte retrieval is more predictive of implantation success than the pattern or thickness on the day of hCG administration. J Assist Reprod Genet 1999;16:523– 8. Kim CH, Chae HD, Huh J, Kang BM, Chang YS, Nam JH. Relationship between endometrial estrogen and progesterone receptors, and sonographic endometrial appearance in the preovulatory phase. J Obstet Gynaecol Res 2000;26:95–101. Scemiczky G, Wramsby H, Johannisson E, Landgren BM. Importance of endometrial quality in women with tubal infertility during a natural menstrual cycle for the outcome of IVF treatment. J Assist Reprod Genet 1998;15:55– 61.
Vol. 78, No. 2, August 2002