SIEP publications are protected by copyright and all rights are reserved.  SIEP publications may not be reproduced in any form or by any means without written permission from the copyright owner. This includes the posting of electronic files on the Internet, transferring electronic files to other persons, distributing printed output, and photocopying. Requests for authorization to make photocopies should be directed to:   SIEP, 1697 Lark Lane, Cherry Hill, NJ 08003-3157
Tel: (856) 429 2699, Fax:  (856) 429 7414, Email: barnea@earlypregnancy.org.

---

EARLY PREGNANCY: Biology and Medicine Editor-in-Chief: Eytan R. Barnea MD, FACOG

July 2000
Volume IV, Number 3
ISSN: 1537-6583
Pages: 191-199

Azadeh Sabi Kaider, BS, Carolyn B. Coulam, MD

The Center for Human Reproduction, 750 N. Orleans St., Chicago, IL 60610, Tel: (312)397-8000 Fax: (312)397-8382

Abstract

The role of luteal phase estrogen in pregnancy outcome has been a matter of considerable debate. In order to evaluate the effectiveness of estrogen supplementation in gonadotropin releasing hormone agonist (GnRHa)/human menopausal gonadotropin (hMG)-stimulated cycles associated with low luteal estrogen concentration, a study was performed comparing the ongoing pregnancy rates in cycles with serum concentrations of estradiol (E₂) <100pg/ml 11 days post embryo transfer (p-ET), treated with luteal phase progesterone (P₄) vs. E₂ and P₄ supplementation. Among 1106 serum samples studied, 951 were from women receiving GnRHa and follicle stimulating hormone (FSH) prior to oocyte retrieval and P₄ (50mg - 100mg IM daily) as luteal phase supplementation beginning day 11 after retrieval. The remaining 155 were from women receiving both E₂ (2mg - 6mg estrace orally each day) and P₄ during the luteal phase. Significantly greater frequencies of preclinical losses were observed among women with human chorionic gonadotropin (hCG) concentrations>5mIU/ml and concurrent E₂ concentrations <100pg/ml compared with E₂ >100pg/ml (p<0.00001). Among the 128 women who had hCG concentrations >5mIU/ml and E₂ concentrations <100pg/ml, 102 received P₄ only during the luteal phase and 26 were treated with estrace 2mg - 6mg daily, as well as P₄ during the luteal phase. The frequency of preclinical pregnancy losses among the 102 women with hCG >5mIU/ml and E₂ <100pg/ml who did not receive luteal E₂ supplementation was 72%, compared with 50% who received luteal E₂ supplementation (p=0.04) The increase in preclinical pregnancy loss rates among women not receiving luteal E₂ resulted in a decrease in ongoing pregnancy rate (8%), compared to those receiving luteal E₂ supplementation (31%) (p=0.002). Our results indicated that a subset of women losing pregnancies preclinically after GnRHa and FSH stimulation due to low luteal phase serum E₂ level may benefit from luteal estrogen supplementation. More sensitive and specific markers are needed to identify prospectively women in this risk group.

Introduction

Materials And Methods

Between January 1, 1994 and July 31, 1995, 1106 serum samples were prospectively drawn from women 11 days after ET. Of the 1106 sera, 951 were from women who had been stimulated with GnRHa (leuprolide acetate, Lupron; TAP Pharmaceutical, North Chicago, IL) and follicle stimulating hormone (FSH) (Metrodin; Serono Laboratories, Randolph, MA) prior to retrieval and P₄ in oil 50mg intramuscularly each day after retrieval of oocytes. The remaining 155 were from women receiving estrogen (Estrace; Mead Johnson Pharmaceutical, Evansville, IN) and P₄ prior to ET.

Each serum sample was evaluated for human chorionic gonadotropin (hCG) and E₂ concentrations. Mean serum E₂ concentrations between women with hCG concentrations greater and less than 5mIU/ml were compared. All cycles in which hCG concentrations were greater than 5mIU/ml were considered indicative of pregnancy and were followed until termination. Pregnancy outcomes were classified as preclinical or clinical loss or ongoing pregnancy (11). A preclinical loss was defined as two rising serum concentrations of hCG without a gestational sac documented by transvaginal ultrasonography three weeks from ET. A clinical loss was defined as an abortion after ultrasonographic documentation of a gestational sac within three weeks of ET using a 5-7.5 MHZ transvaginal probe. Chromosomal analysis was obtained from seven of the products of conception using standard G-banding technique. An ongoing pregnancy demonstrated ultrasonographic evidence of normal progression of the pregnancy from the first to the second trimester.

Women with serum hCG concentrations of greater than 5mIU/ml were divided into two groups based upon their concurrent E₂ concentration: Group 1 had serum E₂ concentrations less than 100pg/ml and Group 2 greater than 100pg/ml. The frequency of preclinical and clinical pregnancy losses were compared between the two groups. Twenty-six women with hCG concentrations greater than 5mIU/ml and concurrent E₂ concentrations less than 100pg/ml were treated with estrace orally (2mg - 6mg) after 11 days from ET. The dosage of estrace was dependent upon what medication the patient was given at the time blood was drawn. If the patient’s E₂ concentration was less than 100pg/ml, estrace dosage was increased by 2mg per day. The frequency of preclinical and clinical pregnancy losses between women receiving estrogen supplementation and those not receiving estrogen during the luteal phase were compared.

Hormone Assays
Blood was drawn between 07:30 to 09:00 hours from the antecubital vein of each woman for determination of E₂ and beta subunit human chorionic gonadotropin (bhCG). Serum E₂ concentrations were determined by radioimmunoassay (RIA; Coat-A-Count, Diagnostic Products, Los Angeles, CA). The intrassay and interassay coefficients of variation were 6% and 5%, respectively. Serum bhCG concentrations were determined by immunoradiometric assay (IRMA; Serono MAIACLONE). The coefficient of variation for bhCG of 25mIU/ml was 5%.

Statistical Analysis
Comparisons between groups were analyzed using Chi-square and Fischer’s Exact tests. In all analyses, p <0.05 was considered statistically significant.

Results

Among 1106 serum samples studied, 951 were from women receiving GnRHa and FSH prior to oocyte retrieval and P₄ as luteal phase supplementation beginning the day after retrieval. The remaining 155 serum samples were from women receiving estrogen during the follicular phase of the cycle and both estrogen and P₄ during the luteal phase. Concentrations of hCG greater than 5mIU/ml were found in 446 serum samples and less than 5mIU/ml in 660 samples. Of the 446 sera with hCG concentrations greater than 5mIU/ml, 366 were from the group of women undergoing stimulation with GnRHa and FSH; and of 660 sera with hCG less than 5mIU/ml, 585 were from GnRHa and FSH stimulated women. Mean E₂ concentrations for women undergoing GnRHa and FSH stimulation who had hCG of greater than 5mIU/ml was 485 + 719pg/ml and hCG of less than 5mIU/ml was 61 + 114pg/ml (p = 0.0001).

One hundred twenty-eight women with serum hCG concentrations greater than 5mIU/ml had concurrent E₂ concentrations of less than 100pg/ml and 318 had E₂ concentrations greater than 100pg/ml. The frequency of preclinical and clinical pregnancy losses as well as ongoing pregnancies within each of the E₂ groups is shown in Table 1. Significantly greater frequencies of preclinical losses and fewer proportions of ongoing pregnancies were observed among the women with hCG concentrations greater than 5mIU/ml and concurrent E₂ concentrations less than 100pg/ml compared with greater than 100pg/ml (p<0.00001).

Of the 128 women with serum concentrations of hCG greater than 5mIU/ml and concurrent E₂ concentrations less than 100pg/ml, 102 received no treatment and 26 were randomly given estrace 2mg - 6mg a day beginning 11 days after ET when the first serum concentration of E₂ of less than 100pg/ml was detected. The frequency of preclinical and clinical pregnancy losses as well as ongoing pregnancies within the estrace treated and not treated groups are shown in Table 2. Fewer preclinical losses (p=0.04) and significantly greater ongoing pregnancies (p=0.002) were observed in the estrace-treated compared with the not treated group. The frequency of clinical losses was the same in the estrace treated and not treated groups. A total of 26 pregnancies were diagnosed as clinical losses (5 receiving estrace and 21 with no treatment). Eight of the clinical pregnancy losses were ectopic pregnancies and 18 were spontaneous abortions. Chromosome analysis was obtained from 12 of the 18 spontaneous abortions. All 12 had abnormal karyotypes: 9 aneuploidy and 3 polyploidy.

A correlation between initial bhCG concentration and pregnancy outcome was observed among the 26 women treated with estrace (Table 3). Mean values for initial bhCG concentrations were lower in those pregnancies subsequently ending in preclinical loss. However, no threshold value for hCG was apparent among the ongoing pregnancies, which ranged from 5.7 to 209mIU/ml. No differences in mean hCG concentrations between women receiving estrace supplementation and no supplementation were observed in pregnancies ending in preclinical loss, clinical loss or ongoing pregnancy (Table 3).

Discussion

Correlations between late luteal serum E₂ concentrations and conception compared with non-conception cycles have previously been reported for both natural (12) and supraovulated (13) cycles. Muasher et al (14) found no statistically significant difference in luteal E₂ concentrations between conception and non-conception cycles until luteal day 11. The single successful pregnancy in an agonadal woman without early estrogen support occurred in a woman who accidentally discontinued all estrogen support on the day preceding ET but was reinstated on luteal day 12 or 9 days after ET (9). Serum concentrations of E₂ during the luteal phase in this patient were not reported. Correlations between luteal E₂ levels and pregnancy outcomes among conception cycles have been published (15). As in the present study, decreased luteal E₂ concentrations have been associated with early pregnancy wastage (15).

That decreased luteal E₂ serum concentrations were associated with unsuccessful pregnancies lead to the question whether luteal phase supplementation of estrogen in addition to P₄ would increase the frequency of ongoing pregnancies after IVF/ET. Two studies have shown no benefit of the addition of E₂ valerate to P₄ luteal phase support of GnRHa/hMG stimulated IVF/ET cycles (6,7). Both of these studies prospectively randomized all patients undergoing ET after IVF to one of two luteal supplementation regimens including P₄ supplementation alone or E₂ and P₄ supplementation. In their discussion of their results, Smitz et al (6) indicated that a potential benefit of E₂ supplementation might be exclusively limited to 25% of the pregnancies in which the first rise in serum hCG concentrations was noted after day 12. They cautioned that a much larger number of patients would be required to detect statistical differences in one quarter of the pregnancies after GnRHa/hMG stimulation (6). The sample size in their study was 378 patients with 139 (37%) patients demonstrating first positive serum hCG values 12 days after ET. The remaining 239 patients had negative serum hCG values and would not be expected to respond to luteal phase hormone supplementation. By selecting a subset of individuals with a positive hCG and decreased serum E₂ concentration (<100pg/ml) 11 days after ET, we were able to show a difference in ongoing pregnancy rate between women treated with late luteal phase estrogen supplementation in addition to P₄ supplementation compared to those without estrogen supplementation. Among women with serum hCG concentrations >5mIU/ml and E₂ <100pg/ml 11 days after ET, estrace supplementation begun 11 days after ET enhanced the rate of ongoing pregnancies significantly (p<0.002). However the absolute difference in ongoing pregnancy rates between the estrace treated and non-treated groups was only 20% suggesting the number of pregnancies with low serum E₂ concentrations 11 days after ET that can be rescued with exogenous estrogen supplementation is low. In fact, even with estrogen supplementation only 30% of pregnancies with a positive hCG at 11 days after ET were ongoing and 70% were lost. The proportion of successful pregnancies (30%) in women with serum E₂ concentrations <100pg/ml and hCG >5mIU/ml 11 days after ET who were supplemented with estrace is significantly lower than the ongoing pregnancy rate of 70% in women with serum E₂ concentrations >100pg/ml and hCG >5mIU/ml 11 days after ET (p<0.00001). The lower ongoing pregnancy rate in women with serum E₂ concentrations <100pg/ml is associated with a higher preclinical pregnancy loss rate compared with serum E₂ concentrations >100pg/ml (p<0.00001). The lower preclinical pregnancy loss rate in women with serum E₂ concentrations <100pg/ml treated with estrace compared with <100pg/ml without estrogen supplementation suggests a role of estrogen in establishing endometrial receptivity late in the luteal phase.

While the role of estrogen during the follicular phase of the menstrual cycle for providing a receptive endometrium capable of establishing and maintaining a pregnancy has been accepted, the role of estrogen during the luteal phase is not clear. Studies to evaluate the role of luteal phase estrogen in establishing a receptive endometrium have assessed the endometrium morphologically by histologic dating of endometrial biopsies and functionally by evaluating pregnancy rates after various assisted reproductive technologies (ART). ART have provided clinical models to increase the understanding of the role of hormones in the development of receptive endometrium. From the donor oocyte model in women with absent ovarian function, estrogen depletion during the luteal phase did not affect the morphologic developmental capacity of the endometrium (16). Nonetheless, exogenous estrogen and P₄ are routinely administered during the luteal phase in the women undergoing ET after oocyte donation (17). While most successful pregnancies after ovum donation have occurred in cycles in which the luteal phase was supplemented by both estrogen and P₄, two pregnancies have been reported after luteal supplementation with P₄ alone in women with absent ovaries (9, 18). In the first report (18), serum bhCG concentration was 11mIU/ml and E₂ was 54pmol/L, 12 days after transfer of two 4-cell embryos. At this time it was determined that the patient had discontinued oral intake of estrogen and was taking P₄ supplementation alone. Estrogen supplementation was not reinstituted. The doubling time for hCG for the following days was 1.3. The serum E₂ concentration increased by increments of 67% to 100% every other day over the next two weeks. The authors interpreted the increasing E₂ concentrations to reflect trophoblastic production of estrogen (18). At 5 weeks of gestation, a transvaginal ultrasound showed an intrauterine gestational sac compatible with gestational age. One week later, a normal embryo with detectable heart beat was documented. During the following week (7 weeks gestation), the patient miscarried. Karyotype of the products of conception showed a duplicated 13/14 Robertsonian translocation. This genetic abnormality was considered the cause of the miscarriage (18). The role of the low circulating serum concentrations of E₂ in effecting the abortion is not known in view of the genetic abnormality. In the second report (9), the error relating to discontinuation of estrogen therapy on the day preceding ET was identified 9 days after the transfer at which time the therapy was reinstated. Therefore, the recipient was without exogenous estrogen treatment from "day 15" to and including "day 26." The serum concentration of bhCG on the day following reinstatement of estrogen supplementation was 207mIU/ml. A twin gestation ensued with delivery of normal male and female infants (9). These two reports (9, 18) and the results of the current study suggest that continued estrogen stimulation to the endometrium is not necessary for implantation but may be helpful in maintenance of early pregnancies. However, studies in subhuman primates suggest estrogen is not necessary for maintenance of pregnancy (2). Recipients of fertilized donated oocytes were ovariectomized on the day of ET and exogenous steroid therapy was initiated. The group that received P₄ only did as well as the group that received both estrogen and P₄. Thus it would appear that luteal phase estrogen supplementation is not necessary for implantation or maintenance of pregnancy as long as P₄ supplementation is continued in agonadal individuals. Even the necessity of P₄ supplementation in the maintenance of pregnancy has been questioned by a case report of a continuing donor oocyte pregnancy in an agonadal woman who accidentally discontinued all forms of exogenous estrogen and P₄ therapy 13 days after ET (19). Thus steroid therapy in the late luteal phase is not necessary for maintenance of all donor oocyte recipient pregnancies but may be helpful in some pregnancies. This concept is supported by the landmark experiments of Csapo (20, 21). These experiments established that while the corpus luteum was essential for pregnancy maintenance until the luteoplacental shift (34 days after conception or 32 days after ET) in the majority of women (20), a subgroup of women had serum P₄ concentrations consistent with earlier placental secretion (21).

By contrast to estrogen and P₄ supplementation in ovum donation cycles, it is common practice to support the luteal phase with P₄ alone following GnRHa and hMG cycles. Morphologic examination of the endometrium after GnRHa and hMG stimulation has shown a high percentage of endometrial biopsies taken 6 - 9 days after hCG administration are out of phase (22). Administration of P₄ alone or P₄ in combination with estrogen during the luteal phase are equally successful in correcting the histologic abnormalities after GnRHa and hMG stimulation. Similarly, administration of P₄ alone or in combination with estrogen during the luteal phase are associated with the same pregnancy rates after GnRHa and hMG stimulation (6,7). Hence, the controversy regarding the role of luteal phase estrogen is apparent. We were able to show that a small subset (perhaps 20%) of women losing pregnancies preclinically after GnRHa and FSH stimulation due to luteal phase defect manifest as low serum E₂ level may benefit from luteal estrogen supplementation. More sensitive and specific markers than serum E₂ concentrations are needed to identify these women prospectively.

For personal use. Only reproduce with permission from SIEP.

References

1. Maslar IA. (1988) The progestational endometrium. Semin Reprod Endocrinol, 6, 115-128.

6. Smitz J, Bourgain C, van Waesberghe L, Camus M, Devroey P, Van Steirteghem AC. (1993) A prospective randomized study on oestradiol valerate supplementation in addition to intravaginal micronized P in buserelin and hMG induced superovulation. Hum Reprod, 8, 40-45.

7. Lewin A, Benshushan A, Mezker E, Yanai N, Schenker J, Goshin R. (1994) The role of estrogen support during the luteal phase of in vitro fertilization - embryo transplant cycles: a comparative study between progesterone along and estrogen and progesterone support. Fertil Steril, 62, 121-125.

12. Laufer N, Navot D, Schenker JG. (1982) The pattern of luteal phase plasma progesterone and estradiol in fertile cycles. Am J Obstet Gynecol, 143, 808-813.

13. Hutchinson-Williams KA, Lunenfeld B, Diamond M, Lavy G, Boyers SP, De Cherney AH. (1989) Human chorionic gonadotropin, estradiol and progesterone profiles in conception and non-conception cycles in an in vitro fertilization program. Fertil Steril, 52, 141-145.

14. Muasher S, Acosta AA, Garcia JE, et al. (1984) Luteal phase serum estradiol and progesterone in in vitro fertilization. Fertil Steril, 41, 838-841.

Table 1

Frequency of preclinical pregnancy loss, clinical pregnancy loss and ongoing pregnancies among 446 women whose serum human chorionic gonadotropin concentration was greater than 5mIU/ml 11 days after embryo transfer.

Table 2

Frequency of preclinical pregnancy loss, clinical pregnancy loss and ongoing pregnancies among 128 women whose serum human chorionic gonadotropin concentration was greater than 5 mIU/ml and estradiol was less than 100 pg/ml 11 days after embryo transfer. Comparison between women receiving luteal phase estrogen supplementation and those not receiving luteal phase estrogen supplementation.

Table 3

Comparison of beta subunit human chorionic gonadotropin (bhCG) serum concentrations (mean±SD) between 26 women receiving luteal phase estrogen supplementation and 102 women not receiving luteal phase estrogen supplementation.