Hopital de Bicetre, Le Kremlin Bicetre, Hopital A. Beclere, Clamart, and Institute de Pathologie et de Cytologie Appliquee, Paris, France

Transcription

1 FERTILITY AND STERILITY Copyright 1993 The American Fertility Society Printed on acid free paper in U.S.A. Clomiphene citrate affects cervical mucus and endometrial morphology independently of the changes in plasma hormonal levels induced by multiple follicular recruitment*t Maria Rebecca Massai, M.D.:j: Dominique de Ziegler, M.D. II Valerie Lesobre, M.D.:j: Christine Bergeron, M.D.~ Rene Frydman, M.D. Philippe Bouchard, M.D.:j: Hopital de Bicetre, Le Kremlin Bicetre, Hopital A. Beclere, Clamart, and Institute de Pathologie et de Cytologie Appliquee, Paris, France Objective: To analyze the effects of clomiphene citrate (CC) on cervical mucus (CM) and endometrial morphology independently of hormonal changes encountered when CC is administered for ovulation induction. Design: Volunteers whose ovarian functions were temporarily suppressed (n = 18) by a longacting GnRH agonist and 6 women of similar age suffering from premature ovarian failure (POF) received E 2 and P. Half of the women also received CC (5 mg/d, days 2 to 6). Setting: Tertiary University Institution, Hopital A. Bechhe. Patients, Participants: Eighteen volunteers suffering from infertility not related to a uterine cause and 6 women of similar age suffering from POF. Main Outcome Measure: Plasma gonadotropins, E 2, and P were measured at baseline to confirm that the ovaries were inactive and twice weekly during physiological E 2 and P replacement. Cervical mucus was analyzed on day 14 and scored from to 15. Endometrial biopsies were obtained on replacement days 2 and 24 for conventional histology and immunocytochemistry analysis of estrogen receptors and progesterone receptors (PR). Premature ovarian failure women whose results have been previously published served as controls for day 2 biopsies. Results: Cervical mucus scored lower in women who received CC (5.5 ± 3.2) than in controls (13.6 ± 4.7, mean± SEM). On day 2, endometrial findings were similar in women treated with CC and in controls. On day 24, specimens showed a significant delay in endometrial maturation in women treated with CC. On day 24, only staining for PR selectively persisted in endometrial stroma, and no difference was observed between women who received CC and controls. Conclusion: Our results indicate that CC significantly alters CM quality and late luteal phase endometrial morphology despite physiological levels of plasma E 2 and P. Hence, clinicians should monitor E 2 levels when using CC, and caution should be exerted when supraphysiologicallevels of E 2 are not present to counterbalance the effects of CC on the CM and the endometrium. Fertil Steril 1993;59: Key Words: Clomiphene citrate, cervical mucus, endometrial morphology, controlled ovarian hyperstimulation Received September 28, 1992; revised and accepted January 28, * Presented at the 38th Annual Meeting of the Society for Gynecologic Investigation, San Antonio, March 2 to 22, t Supported in part by a grant from Ciba Pharmaceuticals, Rueil, France. t Service d'endocrinologie et des Maladies de Ia Reproduction; Hopital de Bicetre. Service de Gynecologie Obstetrique, Hopi tal A. Beclere. Numerous investigators have reported that clomiphene citrate (CC) adversely affects cervical mucus (CM) quality and endometrial morphology (1-4). It II Reprint requests: Dominique de Ziegler, M.D., Department of Obstetrics-Gynecology, Hopital A. Beclere, 157 Rue de Ia Porte de Trivaux, Clamart, France. 1l Institut de Pathologie et Cytologie Appliquee. Massai et al. Effects of CC 1179

2 is not clear, however, whether these findings observed in CC cycles result from an action of CC itself or from the excess, and the possible imbalance, in hormonal levels induced by the multiple follicular recruitment triggered by CC (5). Indeed, even in the absence of CC, the excessively high levels of plasma E 2 encountered in controlled ovarian hyperstimulation (COH) have been blamed for altering endometrial morphology (5, 6). Still another mechanism has been put forth to explain the CM alterations and the abnormal endometrial morphology often seen after CC administration. It has, indeed, been hypothesized that an inadequate follicular maturation can lead to premature luteinization (7) and/ or luteal phase defect (8, 9). Today, most investigators concede that in a fraction of users, CC can harm CM and endometrial morphology. Yet, in spite of the long clinical experience, the mechanisms at play that explain these effects of CC are still debated. To clarify how CC affects CM and endometrial morphology independently of any changes in hormonal levels, we studied the consequences of administrating CC in conjunction with set levels of plasma E 2 and P. For this, women whose ovaries were temporarily suppressed by a long-acting GnRH agonist (GnRH-a) were prescribed 1 mg of CC from days 2 to 6 while also receiving physiological E 2 and P replacement. MATERIALS AND METHODS Subjects' Characteristics Twenty-four women, 31 to 39 years of age, volunteered for the study after giving an informed consent. Of these, 18 women having regular ovulatory cycles were candidates for IVF, having infertility reasons that did not involve a uterine cause. The remaining 6 women suffered from premature ovarian failure (POF) and were candidates for IVF with oocyte donation. Body weights of all the study participants were within 1% of ideal weight for size. None of the participating women received any hormonal treatment within 6 weeks of starting the study. Hormonal Treatment A single intramuscular injection of a timed release preparation of GnRH-a (Decapeptyl-Retard, 3.75 mg; Ipsen-Biotech Pharmaceuticals, Paris, France) was administered on cycle day 2 in 18 women having functioning ovaries. Estradiol and P replacement was initiated 2 or 3 weeks after confirmation of pituitary and ovary desensitization was obtained by measurement of plasma LH and E 2 Thereafter, treatment consisted of.1 to.4 mg of E 2 administered transdermally following a regimen designed to duplicate the plasma E 2 and estrone levels normally observed in the menstrual cycle as reported in women with inactive (1, 11) or temporarily suppressed ovaries (12). Briefly, E 2 was administered from one or several transdermal therapeutic systems (Estraderm-TTS 1; Ciba Pharmaceuticals, Rueil, France) each delivering approximately.1 mg of E 2/d. Moreover, micronized P (Utrogestan; Besins Iscovesco Pharmaceuticals, Paris, France) was administered vaginally (3 mg/d) from days 15 to 28. In addition, 12 ofthese women also received 1 mg of CC (Clomid; Merrel Dow Pharmaceuticals, Neuilly-Sur-Seine, France) orally from cycle days 2 to 6. The remaining 6 women whose ovaries were suppressed by GnRH -a, received E 2 and P only and served as controls (day 24 biopsy group). The 6 POF women whose data have been previously reported (11) received E 2 and P only and served as controls (day 2 biopsy group). Blood Sampling and Hormonal Assay Baseline blood samples were obtained by venipuncture 2 to 3 weeks after pituitary desensitization in women who received GnRH-a and before E 2 and P treatment in POF patients. During treatment, blood samples were obtained two times per week, 24 to 36 hours after the placement of one or several new transdermal systems to minimize the fluctuations in plasma E 2 related to a known imperfection in the pharmacokinetics of transdermal systems (13). Hormonal measurements included the determination of plasma E 2, P, LH, and FSH. Plasma E 2 and P were measured by a double-antibody RIA technique and LH and FSH by an immunoenzymatic technique, as previously reported (11). Intra-assay and interassay precision was in all cases < 11% and 13%, respectively. Cervical Mucus Assessment and Endometrial Sampling Cervical mucus was analyzed on days 13 and 14 and scored from 1 to 15 according to Moghissi (14) by a team member not directly aware of the study. Briefly, mucus quantity, aqueous content, cellularity, ferning, and spinbarkeit were each assessed and scored from to 3. The arithmetical sum served as final mucus score value. An endometrial biopsy was performed using a four-quadrant scraping technique 118 Massai et al. Effects of CC Fertility and Sterility

3 immunoglobulin (lg) (diluted 1:2) (Amersham Pharmaceuticals, Buckinghamshire, United Kingdom) and streptavidin-peroxidase complexes (diluted 1:1) (Amersham). The ER antibody was detected with an anti-rat Ig and rat monoclonal peroxidase-antiperoxidase complex (Abbot). Peroxidase activity was demonstrated by the diaminobenzidine reaction. Duplicate sections were slightly counterstained with hematoxylin to facilitate the identification of cellular elements. A section of each tissue block adjacent to the immunostained section was subjected to a similar treatment, except that the primary mouse anti -PR antibody or rat anti -ER antibody was replaced by a mouse receptor-unrelated mab or rat normal Ig used at the same concentration. The distribution of peroxidase reaction product was examined by light microscopy. The intensity of specific staining of individual cells was characterized as absent (), weak (+), moderate (), or intense (+). The proportion of cells specifically stained was estimated by eye as absent (), 25%, 5%, or 75% of the total cell population. Results were separately scored for the glandular and stromal components of the endometrium after the slides were coded so that the examiners (M.R.M., D.Z., P.B.) were ignorant of the clinical history. All evaluations were made at X25 and X4 magnification. Figure 1 Endometrial biopsies obtained on day 2. Results are similar in women who received CC and in controls. In both groups, findings are characterized by an abundant development of subnuclear vacuoles in the glandular epithelium. These vacuoles, developed at the base of epithelial cells, have pushed the nuclei to a more apical position, thereby staging the characteristic palisade appearance that is typically seen on day 18 of the menstrual cycle. on cycle days 2 and 24 (1 biopsy per patient, n = 6 in each biopsy group). Control women whose ovaries were suppressed with GnRH -a were biopsied on day 24 (n = 6). The control group for day 2 biopsies (n = 6) consisted of POF women who received the same E2 and P treatment. Endometrial specimens were immediately halved in two equal parts. One part was immersed in Bouin (Farmitalia Labs, Rueil, France) for further paraffin embedding and conventional histology analysis. The endometrium was dated according to the criteria of Noyes et al. (15) after coding the slides so that the two investigators, M.R.M. and C.B., were not aware of the clinical treatment received. The second part (day 24 biopsies) was immediately snap-frozen in liquid nitrogen and stored in liquid nitrogen for subsequent immunostaining for estrogen receptors (ER) and progesterone receptors (PR). Briefly, endometrial samples were immunostained on adjacent frozen sections with monoclonal antibodies (mabs) for ER and PR (16). All tissue samples were analyzed in duplicates. Tissue samples were sectioned (5-#Lm thick) at -26 C and were thaw-mounted on gelatincoated glass slides. The anti -PR antibody LET 126 (6 #Lg/mL) was used (16). The details of the preparation of this antibody and its specificity have been described previously (1 7, 18). Immunostaining for ER was performed with a kit obtained from Abbot Pharmaceuticals (North Chicago, IL). The PR antibody was detected with biotinylated anti-mouse Statistics Between -group comparisons were assessed by Student's t-test, making Bonferoni adjustments for multiple comparisons. Probability <.5 was considered as _statistically significant. RESULTS Hormonal Levels Plasma E2 and P levels are illustrated in Figure I 1. Baseline plasma E2 and P levels were not different from findings made in menopausal women using the same assay, thus confirming that GnRH -a achieved complete ovarian suppression. After physiological E2 and P replacement, the profiles of plasma E2 and P levels were similar in women who received CC or not (controls) and were also similar to those previously observed in POF patients not receiving GnRH-a (1,11). Baseline and during treatment levels of plasma LH and FSH, however, differed between women who received GnRH -a and POF patients. In Massai et al. Effects of CC 1181

4 normally cycling women receiving GnRH -a, plasma LH and FSH remained <2.5 IV /L throlj.ghout the observation period. As previously reported, however, baseline LH and FSH levels were in the menopausal range in women whose ovaries were inactive. In these later women, values obtained during treatment showed a progressive decrease in response to E2 and P administration (11). Cervical Mucus Score and Endometrial Biopsies In women receiving E2 and P only (controls) the mean CM score on days 13 and 14 was 13.6 ± 1.2 (mean ± SEM), a value not different from those observed at the end of the follicular phase of the menstrual cycle. Among the controls receiving E2 and P only, there was no difference between the women whose ovaries were temporarily suppressed with GnRH -a and the POF patients. Women who received CC, however, showed a significant decrease in their mean CM score to.5.5 ±.9, mean ± SEM (P <.1). Conventional histology and immunocytochemistry findings are summarized in Table 1. Conventional histology obtained on day 2 showed similar findings in women who received CC and controls. The characteristic aspect of day 2 endometrial morphology observed in women who received CC and in controls is illustrated in Figure 1. It is characterized by early secretory changes with an abundant development of subnuclear vacuoles in the glandular epithelium as typically ' observed on day 18 of the menstrual cycle. Day 24 endometrial biop- Figure 2 Endometrial biopsies obtained on day 24 in treated (A) and control groups (B). In the treated group (A), a delay of at least 2 days is observed in the development of secretory changes. Rounded endometrial glands have nuclei that regained their original basal position while intraluminal secretion of glycogen is intense. The stroma is edematous and no sign of predecidualization can be seen. In the control group (B), spiral arteries are well differentiated and are surrounded by clear signs of predecidualization of stromal cells, as typically seen on day 24 of the menstrual cycle. Table 1 Individual Results of Endometrial Biopsies on Cycle Day 24 PRt ERt Subject no. Dating* Stroma Glands Stroma Glands CCI CC2 CC3 CC4 CC5 CC6 Control 1 Control 2 Control 3 Control 4 Control 5 Control /- NA NA +/ * Histologic dating as per Noyes et al. (15). t Staining for ER and PR was graded as:, absent; +, weak;, moderate; +, intense; NA, not available Massai et al. Effects of CC sies are illustrated in Figure 2A for women who received CC and in Figure 2B for controls. Women who received CC showed a delayed development of secretory changes by comparison with endometrial findings made in controls. Namely, no development of spiral arteries and no predecidual changes of stromal cells were seen in the CC-treated women. In contrast, endometrial specimens obtained on day 24 in controls were similar to findings normally made on day 24 of the menstrual cycle. Immunoanalysis of ER and PR was performed Qn endolnetrial specimens obtained on day 24. There was no difference in ER and PR staining between women who received CC and those who did not. Only staining for PR was seen in the nuclei of stromal cells (Fig. 3), whereas no specific marking for ER Fertility and Sterility

5 A B Figure 3 Immunostaining for PRs on day 24 in the treated (A) and in the control (B) groups. Glands (G) and stroma (8) are visible in all preparations (X25). Progesterone receptor staining is visible in stromal cells only in both groups. could be detected in either glandular or stromal cells (data not shown). This pattern of PR and ER distribution was not different from that reported in the menstrual cycle, starting from day 21 of the menstrual cycle onward. DISCUSSION In the present study, we used the E2 and P replacement cycle model to study the impact of CC on CM and endometrial morphology. The major asset of this paradigm lies in the possibility of analyzing the consequences of a single alteration, here the adjunction of CC, on reproductive parameters without fearing that interferences triggered by feedback mechanisms might confuse the results. As previously reported (12), women who were spontaneously or therapeutically (GnRH -a) deprived of their ovarian function had similar plasma E2 and P levels when receiving physiological E2 and P replacement. The only hormonal difference between women whose ovaries were spontaneously inactive and those who received GnRH -a lay in plasma gonadotropin levels. Before E2 and P treatment, baseline LH and FSH levels were elevated in women whose ovaries were absent and progressively declined thereafter under the influence first, of E2 and later, of E2 and P (11). On the contrary, in women receiving GnRH -a, plasma LH and FSH remained <2.5 lull at all times throughout the study, which precluded studying the effects of CC on plasma gonadotropins. The delay in endometrial maturation that we observed on day 24 in women treated with CC is in keeping with some previous reports made in patients who received CC to induce or augment ovulation (18). Yet our findings further substantiate this phenomenon. They indicate that endometrial alterations observed in the late luteal phase after receiving CC from days 2 to 6 result from an effect of CC itself rather than from an alteration in hormonal levels induced by CC. On the contrary, there are investigators who have failed to observe any morphological abnormality in late luteal phase endometrial specimens from CC-induced COH cycles (19, 2). A hypothetical explanation for this puzzling paradox is that in CC-induced COH cycles, the supraphysiological levels of E2 may counteract the anti-estrogenic properties of CC. In our model, CC interfered with the development of day 24 endometrial morphology in all the specimens examined. However, the distribution and tne intensity of immunospecific staining for ER and PR in glandular and stromal components of the endometrium were not altered by CC. This apparent contradiction between CC altering day 24 endometrial Massai et al. Effects of CC 1183

6 morphology yet not modifying specific immunostaining for ER and PR is not truly paradoxical. Indeed, day 24 endometrial specimens of CC-treated patients were characterized by a developmental delay unveiling a typical day 21 to 22 morphological aspect. But the pattern ER and PR distribution that characterizes day 21 to 22 and day 24 endometrial specimens are similar, embodying the disappearance of ER and PR from endometrial glands and the selective persistence of staining for PR in the nuclei of endometrial stromal cells (16). Therefore, our finding of similar immunocytochemistry results on day 24 in CC-treated patients and controls is perfectly consonant with the physiological pattern of ER and PR distribution seen during the course of the late luteal phase. Hence, these results do not constitute a dissociation between the morphological and the functional (ER and PR distribution) aspect of the endometrium. Contrasting with the delay in the secretory transformations of the endometrium observed on day 24 in women who received CC, endometrial specimens obtained on day 2 did not differ between women who received CC and controls. In both groups, day 2 endometrial specimens displayed findings typically seen on the 4th day after ovulation in the menstrual cycle (day 18). This aspect is characterized by an abundant development of sub nuclear vacuoles in the glandular epithelium pushing the nuclei to a midcellular location. Therefore, by comparison with the menstrual cycle, day 2 endometrial specimens display a 2-day lag in the secretory transformation of the endometrial glands. This maturational delay of the glandular epithelium has been widely recognized by all the investigators who have studied the E 2 and P model. Putative mechanisms explaining this delay in the secretory transformations of endometrial glands have been discussed in details elsewhere (21). In the context of the present study, however, this particularity of E 2 and P cycles does not affect data interpretation. Indeed, in our experimental paradigm the effects of CC were judged by comparing biopsies obtained in women receiving similar treatment except for CC (controls) rather than by paralleling results to theoretical references taken from the menstrual cycle. Hence, the claim that CC does not affect the secretory transformations of day 2 endometrium is valid. A controversy has existed in the literature about the possibility that CC affects the endometrium during the early to midluteal phase, that is, at the theoretical time of embryo implantation. In this respect, our observation that CC does not affect day 2 endometrial specimens agrees with the reports that have failed to observe endometrial effect of CC during the midluteal phase (22 to 23). However, we do not have a definitive rationale for the discrepancy between our results and the reports of altered endometrial morphology propounded by other groups (24). One plausible explanation for these divergent observations is that endometrial alterations reported on day 2 in CC-induced COH cycles (24) might have resulted from a hormonal imbalance induced by CC rather than from CC itself. The reports of altered endometrial morphology on days 18 to 2 in COH cycles not using CC certainly speaks in support of this later concept (5). We cannot readily explain the heterogeneous character of the endometrial effects of CC that affects endometrial morphology on day 24 but not on day 2. A seemingly logical hypothesis for rationalizing this delayed effect of CC on the endometrium proposes that CC interferes with late luteal changes through persistant tissue accumulation. According to this hypothesis, CC accumulated in the endometrial stroma could hinder the action of luteal estrogen (E) on late luteal phase secretory changes of the endometrium (25). Yet our previous observation that interrupting E 2 supply past day 15 in the E 2 and P replacement cycle model did not interfere with a normal day 24 endometrial morphology (11) has seriously challenged a putative role of luteal E 2 in late luteal secretory changes. An alternate hypothetical mechanism that we would like to propose for of explaining the observation that CC only alters late luteal endometrial morphology while not affecting early luteal changes is that endometrial glands and stroma respond differently to CC. According to this postulate, endometrial glands that embody the morphological criteria retained by Noyes et al. (15) for dating human endometrium in the early luteal phase would not be affected by CC. In contrast, endometrial stroma that carries the morphological features characteristic of the late luteal phase would be affected by CC as suggested by Fritz et al. (25). It has been documented that CC inherits its anti-estrogenic property from competing with E 2 for ER (9). Hence, it is tempting to assume that the hampering of late luteal endometrial transformations observed after exposure to CC might also result from an action of CC on hormonal receptors. Yet when the endometrial effects of CC are observed, i.e., during the late luteal phase, staining for ER has already disappeared from both glands and stroma, whereas staining for PR persists in the nuclei of the endometrial stroma cells only (16). Therefore, the 1184 Massai et al. Effects of CC Fertility and Sterility

7 data also raise the possibility that CC interferes with the functional integrity of PR. Future work should aim at testing this later hypothesis. The poor quality of CM characterized by insufficient cervical scores has long been recognized in CC cycles as a factor capable of hampering the results of COH cycles (1, 4). The prevailing theory proposed to explain the effects of CC on the cervix has been that the quality of the CM is altered by local anti-estrogenic properties of CC. In turn, this phenomenon can be reverted, at least in part, by supplying large amounts of exogenous E 2 (1). That CC alters the CM by a direct effect resulting from its anti-estrogenic properties has recently been challenged by Taeny et al. (7). These authors have reported that in a high percentage of CC cycles, the poor CM score often results from premature P elevation rather than from an effect of CC itself (7). Although not disputing the clinical merit of this provocative report, our present observation provides experimental data, suggesting that CC also affects CM directly, independently of any hormonal changes occurring in COH cycles. In conclusion, our results indicate that CC significantly lowers the CM score and delays the morphological changes occurring in the endometrial stroma. Yet CC did not affect the secretory transformations of endometrial glands seen in the early to midluteal phase, that is, at the theoretical time of embryo implantation. It is conceivable, however, that in COH cycles using CC, the supraphysiological levels of plasma E 2 that are often achieved might partially counteract the anti-estrogenic properties of CC. Our results suggest, however, that it is a sound clinical precaution to avoid CC when suboptimal levels of E 2 are feared that may not suffice for opposing the effects of CC on the endometrium and the CM. Acknowledgment. P. Legrain, M.D., Institut Pasteur, Paris, France, is greatly appreciated for providing us with a mouse receptor unrelated antibody. REFERENCES 1. Van Campenhout J, Simard R, Leduc B. Antiestrogenic effect of clomiphene in the human being. Fertil Steril 1968;19: Wentz AC. Endometrial biopsy in the evaluation of infertility. Fertil Steril 198;33: Cook CL, Schroeder JA, Yussman MA, Sanfilippo JS. Induction of luteal phase defect with clomiphene citrate. Am J Obstet Gynecol 1984;149: Randall JM, Templeton A. Cervical mucus score and in vitro sperm mucus interaction in spontaneous and clomiphene citrate cycles. Fertil Steril 1991;56: Forman RG, Eychenne B, Nessmann C, Frydman R, Robel P. Assessing the early luteal phase in in vitro fertilization cycles: relationships between plasma steroids, endometrial receptors and endometrial histology. Fertil Steril 1989;51: Smitz J, Devroey P, Camus M, Deshacht J, Khan I, Staesen C, eta!. The luteal phase and early pregnancy after combined GnRH -agonist/hmg treatment for superovulation in IVF or GIFT. Hum Reprod 1988;3: Taney FH, Grazi RV, Weiss G, Schmidt CL. Detection of premature luteinization with serum progesterone levels at the time of the postcoital test. Fertil Steril 1991;55: Garcia J, Jones GS, Wentz AC. The use of clomiphene citrate. Fertil Steril 1977;28: Birkenfeld A, Beier HM, Shenker JG. The effect of clomiphene citrate on early embryonic development endometrium and implantation. Hum Reprod 1986;1: de Ziegler D, Bessis R, Frydman R. Vascular resistance of uterine arteries: physiological effects of estradiol and progesterone. Fertil Steril 1991;55: de Ziegler D, Bergeron C, Cornel C, Medalie D, Massai MR, Milgrom E, et a!. Effect of E 2 luteal on the secretory transformation of human endometrium. J Clin Endocrinol Metab 1992;74: Schmidt CL, de Ziegler D, Gagliardi CL, Mellon RW, Taney FH, Kuhar MJ, eta!. Transfer of cryopreserved-thawed embryos: the natural cycle versus controlled preparation of the endometrium with gonadotropin-releasing hormone agonist and exogenous estradiol and progesterone (GEEP). Fertil Steril1989;52: Powers MS, Schenkel L, Darey PE, Good WR, Balestra JN, Place VA. Pharmacokinetics and pharmacodynamics of transdermal dosage forms of 17 I)-estradiol: comparison with conventional oral estrogen used for hormone replacement. Am J Obstet Gynecol1985;152: Moghissi KS. Diagnosis and classification of disturbed spermcervical mucus interaction. In: Insler, Lunenfeld B, editors. Infertility: male and female. London: Churchill Livingstone, 1986: Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 195;1: Garcia E, Bouchard P, De Brux J, Berdah J, Frydman R, Perrot-Applanat M. Use of immunocytochemistry of progesterone and estrogen receptors for endometrial dating. J Clin Endocrinol Metab 1988;67: Perrot-Applanat M, Groyer-Picard MT, Lorenzo F, Jolivet A, Vu Hai MT, Pallud C, et a!. Immunocytochemical study with monoclonal antibodies to progesterone receptor in human breast cancer. Cancer Res 1987;47: Logeat F, Pamphile R, Loosfelt H, Jolivet A, Fournier A, Milgrom E. One step immunoaffinity purification of active progesterone receptor. Further evidence in favor of the existence of a single steroid binding subunit. Biochemistry 1985;24: Hecht BR, Bardawil WA, Khan-Dawood S, Dawood MY. Luteal insufficiency: correlation between endometrial dating and integrated progesterone output in clomiphene-induced cycles. Am J Obstet Gynecol199;163: Keenan JA, Herbert CM, Bush JR, Wentz AC. Diagnosis and management of out-of-phase biopsies among patients Massai et al. Effects of CC 1185

8 receiving clomiphene citrate for ovulation induction. Fertil Steril 1989;51: de Ziegler D, Cornel C, Bergeron C, Hazout A, Bouchard P, Frydman R. Controlled preparation of the endometrium with exogenous estradiol and progesterone in women having functioning ovaries. Fertil Steril1991;56: Rogers PAW, Polson D, Murphy CR, Hosie M, Susil B, Leoni M. Correlation of endometrial histology, morphometry, and ultrasound appearance after different stimulation protocols for in vitro fertilization. Fertil Steril1991;55: Lamb FJ, Colliflower WM, Williams JW. Endometrial histology an coception rates after clomiphene citrate. Obstet Gynecol 1972;39: Cook CL, Schroeder JA, Yussman MA, Sanfilippo JS. Induction of luteal phase defect with clomiphene citrate. Am J Obstet Gynecol 1984;149: Fritz MA, Westfahl PK, Graham RL. The effect of luteal phase estrogen antagonism on endometrial development and luteal function in women. J Clin Endocrinol Metab 1987;65: Massai et al. Effects of CC Fertility and Sterility